I've been using Signal for a long time. I have repeatedly been unable to convince iOS users to use Signal because "I don't want another app". Android users have been much more willing to give it a shot.
As an android user myself, I much prefer having SMS built in because I use the search feature often to look back through all my SMS/Signal chats. I also regularly forward an SMS message to a Signal user, or vice versa. I'm already starting to feel like those iOS users who told me "I don't want another app"...
Signal seems to be trying to move further and further from "my preferred way to chat with people" and closer to the chat equivalent of "that protonmail account I only log in to when I need secrecy".
I obviously love having security on messages in transit, but I also like being able to keep my message history around and search my conversations for something that happened a year ago. It seems like Signal is on a trajectory to turn everything into disappearing messages. Are they the "safe for activists" communication app, or the "let's try to make as many as possible safer by default" app? Feels like they don't know.
And on top of it all the messaging is just frustrating. "we've taken away an incredibly useful and heavily used feature so we have development resource to better implement shitcoins and such" is such an irritating defense of the decision that I disabled my monthly donation.
I agree completely here. This is terrible news from my perspective too. I use Signal for _all messaging_ (e2e secure or not) for the reasons that you mention.
I've onboarded friends and family, too, ensuring them it should be set as their default messaging app and that it _just works_. Unfortunately, people in the general population seem to have pretty much zero tolerance for any friction whatsoever. If they have to use 2 apps, they'll just end up communicating with me in the clear using their "default SMS" app on their phone. That's what this is going to result it...a reduction in overall message security due to people defaulting to what's easier...which is to _not_ have to remember which app to use for which "send a message" purpose. Fuck.
I understand the argument about people in markets where SMS is expensive getting screwed sometimes when they don't realize they're sending a message over SMS. However could that not be fairly trivially solved for with some UI notification or app setting that warns you about this and allows the warning to be perm-disabled if the user doesn't care!?
I think the real reason here is this desire to transition the service into supporting usernames, which is a topic that's been discussed before (and is explicitly mentioned in the post). Right now the service is tied to your phone number. After this change I suspect it will not be or not need to be.
This is very, very unfortunate for those of us who've convinced a ton of non-technical friends and family to use TextSecure->Signal over the years...
If they have to use 2 apps, they'll just end up communicating with me in the clear using their "default SMS" app on their phone. That's what this is going to result it...a reduction in overall message security due to people defaulting to what's easier...which is to _not_ have to remember which app to use for which "send a message" purpose. Fuck.
Exactly right. In fact, I will be doing this too. Right now, I use Signal as my default messaging app, and messages go via Signal to people who've registered their phone number with Signal without me having to do anything to effect that. When Signal stops working as an SMS client, I'm not going to look up which of my contacts wants to use Signal and which can only use SMS - I'm just going to use my default SMS app for everything except the couple of group chats which are on Signal already. Messages which would have been opportunistically encrypted are now going to go out plaintext, because the friction to find out who uses Signal and who doesn't is too high.
This already happens with iOS, because the Signal app doesn't work as a default SMS client. People who use Signal and Android reliably send me messages over Signal, whereas my contacts who use Signal and iOS mostly send messages over SMS unless they're directly replying to something I've sent them over Signal.
And even if I wanted to do the work to send everything over Signal to those contacts, how am I supposed to remember exactly which of my kid's friend's parents use Signal and which don't? Fuck it, the juice isn't worth the squeeze, everything is going over SMS.
I'll just have to hope that Google figures out a way integrate a seamless, opportunistic end-to-end encryption protocol into the default Android messaging client, since Signal is deliberately dropping the ball here.
Holy crap, this 100%. I am absolutely distraught. I give monthly donations, have moved mountains to get family and friends over to Signal, and viewed it as a last bastion for security and interoperability between iOS and Android. Now we will get neither.
Gutted. I will end up switching to a different provider. Absolutely don't know if I'll be able to convince family and friends to switch again.
I don't really see Signal as a useful tool for my uses after this change. I liked the SMS and encrypted messages in one place thing even though some friends rarely check it. For those friends I send SMS through Signal. There are certainly issues with group chats and media but nothing bad enough to consider using two apps. Now I have to look for a new tool, and figure out what to do with certain archive conversations I'd like to retain.
>viewed it as a last bastion for security and interoperability between iOS and Android
>Now we will get neither
Correct me if I'm wrong, but:
- If you and your counterparty are both using Signal, nothing about the security of your communication changes.
- If your counterparty was not using Signal, your messages were over SMS/MMS and therefore not encrypted in the first place, and so once again, nothing about the security of your communication changes.
That's definitely true! But if both are in one app you don't need to remember who uses Signal and who doesn't; it just works. Now with two apps, you have to run that mental check before sending a message and choose the right app.
> I'll just have to hope that Google figures out a way integrate a seamless, opportunistic end-to-end encryption protocol into the default Android messaging client, since Signal is deliberately dropping the ball here.
This is what Google's RCS implementation is, to my understanding.
Although the end-to-end bit might be a bit on the squishy "kinda-maybe-sorta" side, rather than Signal's enforcing of it.
I have (iOS-using) friends I have convinced to register their numbers with Signal...but they never check Signal. They do check SMS.
This actually drove me away from using Signal for SMS on Android, because I couldn't send those friends SMS messages through Signal. They could only go out as Signal messages, and my friends never checked them.
Fortunately, I only have a handful of friends in this category, and virtually everyone else I talk to does use and check Signal, but it's a usability problem both ways.
On my wifes phone she didn't receive notifications for signal because the OS she had would suspend signal when it was inactive. So I had to manually change its battery saving settings.
I use both Signal and SMS on my phone (iOS) and don’t find it to be that cumbersome to deal with. I already have to use multiple apps anyway to communicate with everyone I need to (Slack, some people are only on Telegram, WhatsApp, WeChat so I have to have those apps too).
I default to SMS but if it’s a conversation where I care about privacy I’ll use Signal.
> Right now, I use Signal as my default messaging app, and messages go via Signal to people who've registered their phone number with Signal without me having to do anything to effect that
This is actually a complaint of mine. I've regularly had Signal send secure messages to people who have tried Signal in the past, but have since uninstalled the app (or in iOS's case, they 'offload' apps when you don't use them frequently which silences notifications). This results in me thinking my messages are sent only to find out that they'll never read them. I then get frustrated and have to go back and re-send the message via SMS.
I don't like some of the decisions taken by Signal.
However, if "dropping support for SMS messaging also frees up our capacity to build new features (yes, like usernames)", I think it is something I would not miss.
Besides, I agree with them on the point that SMS leak metadata.
But that's just a bullshit promise (in the sense of not being anchored to any commitment), no different from a politician saying some policy initiative will create jobs because that's how the economy works.
SMS does leak metadata but guess what, that leakage is going to continue because people aren't going to just cut off their SMS-only-using friends and relatives. Now they'll be leaking Metadata from an even less secure app, so the user is in no way better off.
There's another reason for iOS users to avoid Signal: It eats up Gigabytes of storage space, refuses to ever clear it, and the devs are rather resistant to accepting that it's even a problem at all: https://github.com/signalapp/Signal-iOS/issues/4916
Since there’s no server storing media, consider saving the stuff you want and dumping all the old photos and video you don’t. It’s a more secure communication tool, not an archive.
I'd take a decent "export my chats" option. I have chat history that goes back years that it's often convenient to be able to search. I'd love to be able to move it off the device, but instead the Signal backup just keeps getting larger and larger.
To be clear, Signal allows you to backup and restore back into Signal on android, which is great. What I meant is that it would be helpful to be able to export that content out of signal and keep an accessible searchable archive off of the device.
If I set a conversation length limit though it prevents my ability to search back through my history, which is the feature I'd like to preserve. I value that history, it's useful to me.
My ideal solution would be to export any message older than a month to an archive on my NAS, ideally in a format that the app could search on request. Keep my history, keep the on-device space nice and small.
I take advantage of the Android backup feature, and the backup syncs over to my NAS via SyncThing automatically, but that's only useful for restoring a brand new phone up to the latest state.
> My ideal solution would be to export any message older than a month to an archive on my NAS
I'm confused at what is stopping you from doing this?
> ideally in a format that the app could search on request.
Are you not able to import these backups into the desktop client? IIRC it is just reading from a file structure. I don't see why a small script couldn't resolve this. Obviously you wouldn't be able to search on your phone, but you said you didn't want that data on your phone anyways. If you did want to search on your phone from your computer's storage, I think you're asking way too much of them (and in danger of asking them to store data for you, which they never will do). But this is hacker news, and I don't see why you can't hack together that tool in a weekend. Probably just a few beers on a Friday is enough for it tbh.
> I'm confused at what is stopping you from doing this?
As far as I know the backup is encrypted.
> Are you not able to import these backups into the desktop client?
No, the desktop client is not standalone, and ONLY syncs with the a phone to get content. Moreover, if you don't use the client for a period (2-3 weeks in my experience?) it de-syncs that desktop client. Re-connecting that desktop to your phone will only sync messages starting as-of the connection, so there's no way to get the desktop app to pull your whole history. (This is another gripe I have about their sacrifice of actual usability for security that only helps a few very specific use-cases.)
You're probably rigth about hacking something together. Someone has created a library [0] allegedly for decoding the backup files. Friday night is D&D night though, so I haven't had the chance. :-)
I was more thinking to port chats from your phone to the desktop. The desktop client is on the NAS and serves as the "storage" where you can search your full backup. I would guess there would be a more difficult solution where you could hack the phone app to sync all data instead of just when your desktop connects. So that solution might take more than just a Saturday night ;)
I agree that there aren't really any great solutions. But the alternative is that Signal stores your data. I think this would make a lot of users very upset and compromise part of Signal's core mission.
This is one area where, and I know what you're going to say, but Matrix actually does really well. My data, on my server, in my house, not tied to any specific single device I own but distributed across several and backed up in an encrypted manner.
But you've got me thinking about the data export option. Saturday is coming. Hmm...
> and I know what you're going to say, but Matrix actually does really well.
I'm actually a fan of Matrix. I just see Matrix and Signal as different tools. You're probably aware of it, but if not, there is a Matrix bridge for Signal.
> But you've got me thinking about the data export option. Saturday is coming. Hmm...
Hey, I'm really hoping it works. Let me know how it goes!
Sorry, I'm not trying to be rude. But I am confused at what they want. We have continued the discussion and the picture is clearer to me. Though I'm not sure exactly how Signal can help with it. It still appears to me that the user wants both reduced storage but to maintain search history, which are at odds with one another. Unless they expect Signal to store their history, which those expectations should be shot down because that is against their core philosophy. I did suggest a hack that might fit their needs (full history on desktop but not phone).
But I think that's what godelski is saying is a bit of a contradiction.
If this "accessible searchable archive" is stored "off of the device", how exactly would the device access and search the archive? Facebook can store your Facebook Chat history on their servers no problem, but E2E encryption makes this much more complicated for Signal!
It's not impossible, Apple does it with iMessage, although that was actually a pretty recent addition despite all of Apple's resources!
> But I think that's what godelski is saying is a bit of a contradiction.
Exactly. I'm saying you can't have both low storage and full chat history. These two cannot work without a storage server. I have suggested a method of using the desktop client as the storage server but I agree that there is no great solution.
> It's not impossible, Apple does it with iMessage
That does come with significant tradeoffs though. Now we have to trust Apple that they haven't taken our keys and that they still aren't handing over this data. It is a tricky situation.
> There is no way within Signal to just export a whole conversation to a file.
You can on Android and Desktop. On Mac see ~/Library/Application\ Support/Signal for your data. The only issue is with iOS which I agree is an issue, but does not seem to be the parent's issue.
No you can't. You can find that info at the OS level, or if you use the beta you could get at it via the Developer tools (since it's just a specialized browser window), but there's no userland option to export your conversation in the application.
That's why I wrote 'within Signal.'
As a simple example of how this could get messy, suppose I wanted to back up all my Signal messages, install a different operating system, install Signal there, and load up my old messages. I do have the IT skills to just find the database file and move it manually, but I'm only confident about that because I know its an SQLite database, where the keys are stored etc.
It's still an IT problem I would prefer not to have, vs the simple option of 'Do you want to export this conversation, with all the security implications Y/N' or 'Import this previously archived conversation?'
Except that they make it a pain in the ass to save the stuff you want, because there isn't an easy 'export conversation' function. You can archive conversations so that old ones don't clutter up your chat list, but the only thing you can do with an archived conversation is...un-archive it. It has literally zero utility.
Your only way of saving things is to either manually save every picture and video, or manually highlight and copy the text in your conversations. The latter defeats any security arguments (other than of inconvenience) but also throws away useful information like timestamps of messages.
My Signal database takes up many GB on my phone, and it's constantly complaining about running out of space. Much of this is the years-long record of conversations with my wife. I'd like to back these up, but I can't. Are you gonna tell me that I shouldn't be using a secure messaging app to communicate with my own family members?
Precisely! So just give me an option to export my stuff to a password-protected archive. Make me check a "yes I'm fully aware that this means I am now responsible for the security of this information" if you want. But Signal doesn't let me do this.
Telegram has huge amounts of video and audio data but they allow you to easily select and remove the cache for certain chats or sections of chats. It's a UX decision that Signal could "easily" copy.
Signal is in a weird place where they can do no right by users. It's a team of like 25 developers building extremely complex software criticized by people that don't understand security and trivialize everything. Reddit has a lot of evangelists that can't even program. Their community forums are a dumpster fire where users act like "my way or the world is going to end" (see the current username discussion. Most people are fair but you see[0]). Anything they say on Twitter gets spammed with questions about usernames by people that can't be bothered to see that it is in alpha testing and available for custom builds. And on HN everyone criticizes Signal and compares it to Matrix which is always better for every single purpose.
I do like Signal and I think they have done a lot of good. I do think they have a lot of valid criticism against them but also I think a lot of people aren't providing useful criticism (it is a shame that's happening here, on a forum that should be filled with tech experts). People also aren't realistic. A 25 person team working at a non-profit aren't going to have the same development capacity as a 250 person team.
What do you "Matrix is always better for every single purpose"? Are you saying that you really believe that, or characterizing others as saying that wrongly?
I don't know much about either, but I thought both had somewhat new (less-tested) encryption algos (one is 'double-ratchet' or something? that recently has shown security vulnerabilities?)
I'm saying that people put Matrix/Element in competition with Signal. These used to be dominating voices here. I do think the Matrix == Slack and Signal == Text philosophy has become more prominent now (the philosophy I prescribe to). But there are also major discussions about decentralization and users would suggest Matrix was more secure because of that even though at the time group chats were not encrypted (they are now) and E2EE was not enabled by default.
These are purely my observations of the discourse around Signal and should not be taken as a universal truth. Only my subjective reality.
I'm not aware of any major vulnerabilities in Matrix (but I'm not following) closely. I'm also not aware of any in Signal, which I know is frequently audited. There is an SGX attack, but it is often blown out of proportion (highly technical attack that requires an unlocked phone to be in the physical hands of the attacker).
Pretending like Signal is the second coming of Christ and implying that Telegram and all else is insecure and for dumb idiots and may have had something to do with it
That's far from what is happening. Everyone is complaining about Signal, including me.
Telegram gets a lot of shit because they prop themselves up as a privacy app but aren't. The default is that things are not encrypted. They use a proprietary encryption scheme. They store user data on their servers. These are not the marks of a privacy app.
Thanks for letting me know what I think? And we're all aware of Telegram shortcomings here, it was an example and correct me if im wrong but this thread is about Signal?
I explained how imo Signal has burned the goodwill of a some of its early adopters. Meaning not recommending it to my friends and buying my mom an iPhone. Now there's an argument to be made that it's not very smart to dismiss the app, but that's what happened.
I'd be very surprised if they manage to salvage its image at this point
Doing so comes at a cost to privacy — by signal having a hosting server, even if the contents are E2EE, retrieving and storing these contents creates a metadata trail. I actually go over these drawbacks and tradeoffs in a recent blog post: https://cassieheart.substack.com/p/notes-on-e2ee
Who said anything about a hosting server? Why isn't there a simple option to export a conversation to local storage, encrypted or unencrypted, along with a warning that 'your conversation is now leaving the secure Signal zone.'
If you don't trust the other end then disappearing messages should be used, simple.
This is one of the problems with Signal having a bit of confusion about what exactly it's use-case is. There are plenty of cases where locking down the ability to save/view/export messages are valuable, and Signal provides tools to be able to do that. Making that the mandatory case though means that it's harder to adopt as a general-purpose communication platform.
The need to decide if the goal is still to get as many people off of SMS/facebook-messenger as possible, or if the goal is to provide extreme security to dissidents and protestors, or if they're going to spend the effort to be able to do both effectively and let you choose which conversations or messages get which level of protection.
But I could just screenshot all of it, and you'd never know. There's a setting to prevent screenshots, but no way to tell if other persons have it enabled or when their setting change.
It's like when Signal asks you to put in your PIN, but it's the same PIN you use to unlock your phone. There's a different Signal PIN, but that functions as a check for when you install Signal on a new device. Your regular PIN is just a repeat of your phone PIN, and thus adds exactly zero security.
For starters, signal retains a conversation for the length of time you grant. That can be indefinite. The way it is retained is in a local storage database. It is intentionally guarded against export (although this is somewhat unavoidable with backup features on phones), so as to avoid companies like Cellebrite making it easy for LE to overstep their bounds and pull the message database when they take your phone. If you want some kind of export interface, your best option is a screenshot — signal does not take any action that threatens the mutual security between parties as explicitly agreed.
I don't want to be handcuffed for my own security, thanks. An export facility has little bearing on the integrity of my communications since a counterparty could simply have their phone taken over if they forgot to set a PIN or gave the PIN up to law enforcement. A warning that the exported data can no longer be considered secure (just like when I save a file or photo) is sufficient.
Meanwhile if my counterparty and I have communications in Signal that we want to preserve at scale, it's impossible to do so. A simple example would be that I have years' worth of conversations with my wife that I can't easily back up any where. We could export every picture and video by hand and screenshot or copy the text of every message, but that would be extremely time-consuming and tedious.
That's the basic problem. The data is already only semi-secure in that it's subject to exfiltration without consent. It's just inconvenient. And where parties do consent, it's very inconvenient because it's many many hours of work; in the end, it's just more security theater.
The biggest issue I can see with using two separate apps is checking who's on Signal and who isn't. That means opening up Signal to see if they're on there and then switching to SMS if they aren't. I much prefer having both types of contacts in the same UI and it's been obvious to me which messages are secure. Also, when someone then joins Signal, subsequent messages to them automatically get upgraded to being secure with no effort on my part.
Yes... Unless it's able to somehow alert you that you're texting someone with Signal, it seems like Signal will be phased out because everyone will default to SMS, unless they have a reason to use Signal for a conversation, which hurts the entire privacy ecosystem.
There was a time when contacts that use Signal were listed above contacts that are only reachable through SMS.
Those two groups of contacts then got mixed making it much harder to see who's on Signal and who isn't.
Back then I thought that was deliberate. It would be surprising to me, if the desire to avoid that self-inflicted confusion would have contributed to the decision to discontinue SMS support.
The Google Messages app is pretty good. It's a little thing, but it's the only app that supports tapbacks from iOS -- so on the (many) group threads I'm on that have iPhone users, I can see loves/like reactions instead of a flood of texts that say "Jane Doe loved an image".
rcs support works well, the emoji reactions are good. the web ui for it is pretty alright. I use the quick responses and scheduled send from time to time. and it cleans up my 2fa codes automatically.
Also, it now sends emoji reactions over sms which is a nice little graceful degredation from sms.
It's terrible. Every once in a while they decide to convince me that I really don't want to be sending an SMS but I want to use google's messenger of the month, that will inevitably be gone next month.
This is what I'm switching to. I would have suggested Textra, but when I downloaded it just now, TrackerControl told me there are a bunch of tracking libraries in it. Simple SMS Messenger, on the other hand, not only doesn't have any tracking libraries, but doesn't ask for internet access!
I've used other tools in the Simple suite and I love them.
Although I like using F-Droid, this app is also available on the regular Play Store as well.
Update: I switched back to Textra because Simple SMS lacked a way to see timestamps for arbitrary messages. It will show occasional times when a message was received with some lull between it and the previous message. Otherwise, the timestamps are a mystery. I don't like that.
Silence is like a less polished version of Signal. The only important feature I really think it's lacking is a search function. You can export your texts to an XML file though, so to find something from a long time ago I just export to a file and use grep to search through that.
Yeah, I like Chomp. As I recall, I first installed it because it had an optional emoji pack with the old blob emojis in it, which I was miffed to have had taken away from me in an Android update.
The provide a fairly clear rationale beyond the ideological concern: users don't know/see who is on SMS and who isn't, and are being hit with high fees, and they are concerned that users may believe they have privacy when they do not.
These are reasonable issues and concerns, so I don't follow why you would question all of the other decisions they make.
Couldn’t they just do something like have the sms messages in a different color than the rest? Similar to how iOS used blue for iMessage and green for texts? Signal could use an annoying red color for sms to make it even more clear.
If only there were a company that had, at some point, demonstrated that you can use indicators such as colour to indicate to users whether their messages were being transported via SMS or an E2E secure transport layer.
As if Signal couldn't change the color of the text bubbles to some shade of, let's just say, green, to indicate when a user is chatting via SMS instead of a Signal secured message.
No such thing as bad PR, unless you're removing features. Having people green-shame just means there are other people that are rolling their eyes and still getting their name out there.
so put SMS in different tab and let users turn on/off this feature in settings, done, problem solved, but I understand giving users options nowadays ain't trendy
Every friend using iOS that I've convinced to use Signal has uninstalled it. They stay registered, though, so I have to notice that my messages aren't reaching them, and re-send as SMS.
> I have repeatedly been unable to convince iOS users to use Signal because "I don't want another app".
Here's how I've convinced my iPhone friends. I tell them if they actually want to send pictures and videos to me that aren't potato quality they can either switch to an Android, email me, or use Signal. At this point Signal is more like a cross platform iMessage. This tends to move people over because Apple's walled garden makes group chats infeasible with mixed devices.
Signal is another walled garden, where you have no say in their decisions and simply must obey or leave. Consider Matrix if you want to have the freedom.
I have got all my friends / family using signal, but everyone else I know and everyone else they know are using SMS or whatever.
They will no doubt stop checking their signal messages from me same as iPhone users often do and I will have to change to insecure messages for all my messaging.
This will break their trust in my recommendation, as much as it will break my trust in signal.
I get the reasoning over this, but I think they fail to realise that interoperability is the most important feature. I will have to stop donating to signal over this change, it will be useless to me...
When I switched from Android to iOS this was the number one technical regression, and for years my go-to example of nice things Apple keeps from us. It’s unbelievable how Signal is sabotaging itself here.
Just use iMessage. It's not as secure as Signal since the server has the keys, but it's the easiest way to ensure that the majority of text messages you send are encrypted.
There was just an article that said that 88% of teens have an iPhone. That means that almost all of their communication is encrypted.
It's various levels of encryption that are acceptable depending on your risk level.
I'm mainly concerned about SMS spoofing and mass surveillance. iMessage protects against that. The only way the government can read your messages is by serving Apple with a warrant to obtain your iCloud backup.
If I had a lower risk tolerance, I would disable iCloud backups to improve my security.
I don't think people are only concerned about the government, but rather the corporations that own your data (via storing the encryption key to use whenever they want, to look through whatever they want)
In the end signal and imessage (with backups off) provide roughly the same claimed guarantee here (on a technical level at least). They both promise not to secretly inject additional client keys into your account, (to at least obtain a forward-operating 'wire tap'), but the software you're running is all opaque and could certainly do so if they wanted (or were compelled) to.
fwiw, iMessage is actually E2E encrypted (without Apple storing the key) if you either don't have iCloud backup enabled OR don't enable "Message in the Cloud".
Historically WhatsApp filled in where SMS was harder to use, and places tended to use either WhatsApp or SMS ubiquitously. So the communities most affected by Signal dropping SMS don’t have an existing social need for WhatsApp.
> I have repeatedly been unable to convince iOS users to use Signal
We don't seem to have this problem so bad in the UK/Europe. Most people I know have WhatsApp and/or Telegram, and FB messenger, and Signal (in my friend circles); all alongside SMS. I have very few iMessage groups, and use it mostly for 1to1 SMS with people I don't know well.
I haven't met a single person using SMS for so long that I can't even remember the phone I had last time I used it. I do work with a lot different contractors.
> but I also like being able to keep my message history around and search my conversations for something that happened a year ago
I'm dreading the day where I lose or break my iPhone and thus lose my Signal chat history. I've been using Signal for years, but this makes me still prefer other messengers when starting new chats, even if Telegram (for example) is not e2e encrypted.
(Unfortunately, I've long overcome the "I don't want another app" thing and just installed all the messaging apps. Although I kind of want to look into setting up Matrix bridges and merging everything into Element.)
The only reason this isn't a dealbreaker for me is because their sms implementation was so buggy and feature-poor that I would never have used it.
(The only reason I use signal is to talk to my girlfriend. The only reason we use it is early in our relationship I was going through a phase where I adopted annoying privacy tools. I wanted to abandon it, but after years of using it she's developed positive emotional associations between our relationship and signal, so for non-technical reasons she likes to keep using it just to talk to me)
I used to use WhatsApp with my girlfriend because I had awful cell service at work but did have wifi. We switched to Signal and have been using it for years. She only talks to me on it and iMessage for everything else but I use it for SMS too. We would literally get into arguments sometimes because she wouldn't see SMS messages I'd send her with my one bar of 3G or they would be out of order and cause confusion. I love having all of my conversations in one app. I really like Signal and if this change goes through, it might make me seriously consider just getting an iPhone next time.
> The only reason this isn't a dealbreaker for me is because their sms implementation was so buggy and feature-poor that I would never have used it.
It took me a long time to find this comment. The number of SMS messages I never received was ridiculous. I still use the app all the time, but would never use the integrated SMS messages ever again.
I honestly don't remember when I sent or received the last SMS that was not the telco's "Welcome to $country, the rates are ...". I.e. I don't have a strong opinion on this, I simply do not use SMS.
This "I don't want another app" thing is senseless to me. Why? What does it hurt having more than one communication channel? In my experience people that say this generally have no space on their phone, usually because of an unfettered willingness to install the taco bell app and the Starbucks app and whatever else.
Their underlying reasoning is correct. SMS sucks, really really bad. They're a secure communications channel. People see signal and think they're secure. Signal has no business supporting SMS. The on boarding has reached critical mass, the neyworke effect is here. If you're smart you'll abandon SMS altogether forever and just tell people to reach you some other way, not ditch the actual, over the internet encrypted channel.
Maybe in the circles you move, saying "Sorry, I don't SMS, you'll have to download this other app to message me" would fly, but I for one have to interact with a lot of people for which it would not.
I have SMS, right, you kind of have to have a phone number, but to me it's legacy support, I pay for the mobile data. I'll do SMS as a sort of fallback.
The people you want to send you SMS like companies that need your phone number can't do SMS, and the ones you'd rather message you elsewhere sometimes insist on SMS. It's annoying.
This is an awful decision. I've converted some friends and family to Signal over the past years (it took a while) and it is now their default messaging app on their phones. This is going to confuse them and is going to make it difficult for me to keep convincing them that Signal is the route to use. ("Why do I need 3 apps (Android Messages, Signal, Whatsapp) to talk to people?")
I absolutely agree. Personally, I've managed to convert around 3 times as many Android users as iOS users, because of this feature. And the few people who stopped using Signal after starting using it did so because of limitations in the SMS/MMS features (fewer number of users allowed in group text, etc). I fully expect to loose 2/3 of my Signal contacts as a result of this decision, and may drop it myself if the number remaining is too small to be worth running a separate app, as most of the ones left will probably be on Matrix as well.
It also puts a spot-light on the "your phone number is your username" policy. This made perfect sense when you are using Signal for opportunistic encryption of texting. It is much less justifiable when using it as a Silo'd app. I really hope they change that and give people who were waiting for that change time to join before killing SMS support.
> It also puts a spot-light on the "your phone number is your username" policy.
I'm willing to bet that this decision is just jumping the gun by a month or two since usernames are around the corner (code exists, just not enabled. Can be used if built from source).
Though I haven't had a hard time converting (Android) users by using another app. Especially people that already use WA. The "other app" just comes off as normal. Apple is a different ball game because the walled garden, but that's also the weakness because you can't send photos/videos in group chats with mixed devices (but Signal can).
They say “opportunistic” as in similar to how iMessage works. If you’re both on the platform, it’s encrypted but you still can communicate with everyone else from one app.
That’s a major boost for those that might not particularly care about encryption to look for specific messaging apps, while still helping by building out the network slowly over time.
The downside is that they will opportunistically send your messages via Signal. If the recipient chooses to not have SIgnal installed any longer, then your messages go into a black hole.
This became much more of a problem for me after they rolled out their shitcoin; suddenly my techie friends were just not responding to messages, and Signal as my main SMS app was not falling back to SMS for these folks.
Apple has the same problem, and an article and entire process for disabling it out of band, plus a heartbeat so it’s done automatically after a while if you don’t reset your phone. It’s a major problem.
I’ve only done the switch from iOS to Android once and I remember it was a pain for a few days until everything realized I didn’t have iMessage anymore.
Even without an iPhone I sometimes miss texts from people using iMessage because my only occasionally used MacBook seems to randomly like to turn messages back on, and so anything from an Apple user ends up there instead of on my phone. It stays that way until I figure out I’m missing texts and go find them on the MacBook and have to manually turn off messages to it again.
> If the recipient chooses to not have SIgnal installed any longer, then your messages go into a black hole.
For two weeks, messages will be shown as sent but not delivered, and after two weeks Signal will not let you send messages to that number until it reconnects to the Signal servers.
For comparison, Apple automatically sends all SMS messages via iMessage opportunistically, and if the user then switches to another phone, all SMS messages from iOS users will be silently discarded in perpetuity. This is a big problem because the recipient has no idea that they're missing messages, and also if they no longer have access to an iPhone, there's no way for them to deregister their phone number from iMessage.
They will also deregister you automatically after some period of time. What you described is the situation several years ago, but it's much better now.
That's a link to deregister a phone number from iMessage without an iPhone, which is good, but I don't see any text on that page that confirms that they'll deregister you automatically, or if there's any user-visible indication of the issue. If that's the case, then I'm glad they finally addressed it, because it was definitely a problem for far too long.
In that case, Signal's current behavior would be comparable to Apple's, if Apple also deregisters you after a period of inactivity.
> The downside is that they will opportunistically send your messages via Signal. If the recipient chooses to not have SIgnal installed any longer, then your messages go into a black hole.
The user cannot just log out of Signal and have the app on other people's devices automatically fall back on SMS the way it works with iMessage?
A lot of people will just delete an app and think there were no side-effects. There was an article here a few weeks ago about people not cancelling in-app subscriptions after deleting an app. Apple will remind you after it deletes it, Google does not.
Logging out might not even be enough, depending on the logic on Signal’s side. Do they use active devices, or just that an account exists?
Yes, exactly. The ability to send SMS from the Signal app has meant I've been pretty successful in getting Android users to switch to Signal. Every iOS user I know always just goes back to using iMessage. Now many of those Android users won't bother either.
I hope it's communicated well to users who aren't readers of Signal's blog. I have relatives who use Signal, and they rely on its fallback-to-SMS feature, possibly without fully understanding it. I'll make sure they understand and are aware of this change, but others may be in the same position.
That ship sailed a long time ago. Signal's userbase overwhelmingly objected to having their sensitive information permanently stored in the cloud too, but signal ignored them.
This is an awful decision. I've converted some friends and family to Signal over the past years (it took a while) and it is now their default messaging app on their phones. This is going to confuse them and is going to make it difficult for me to keep convincing them that Signal is the route to use.
I learned to stop trying to improve the technical lives of other people after Dropbox's decision to restrict free accounts to three devices resulted in a shitstorm of angry and confused messages from half the people I know.
You know, I haven’t really thought of it like this. Those for whom I take an interest in their technical lives typically get a spiel from me about whatever solution I’m offering. That spiel often includes something about how “they’ll probably change this eventually in ways no one wants, but the most we can do is speak up. We probably won’t get options.”
But I have to admit your perspective calls to me. I can imagine it would feel quite freeing.
I’m in a minor mess of a situation with my dad’s phone and computer because I’ve tried to be helpful. Now he resists help and that makes both of us frustrated.
I'm happy to share the best information I have with others and most of them are glad that I do.
I had recommended signal to others, but thankfully I've already warned those same people against continuing to use Signal years ago. Nobody was mad at me for Signal's actions and changing your default SMS app isn't hard anyway.
I don't think you have to stop recommending things to people just because situations change. Hasn't everybody had some service or software they depended on go from great to shitty? It's just the nature of using someone else's stuff. At some point they get greedy or busy or decide to pivot into something different from what you want and you have to find something new. Isn't everyone used to that? Why would they blame you?
I have never understood why decisions and options have to be mutually exclusive. Yes, you want to have a rock-solid, thoughtfully-design default install for new and casual users. You can still have an advanced control panel with everything a power user could want.
KDE is such a great example of how to do it right that I didn't even think of it. It just works so well and so transparently that I forget how great it is sometimes.
Rock solid and both works and looks great right out of the box. So customizable that using literally anything else feels like using a Fisher-Price computer for toddlers.
That review of KDE is so over-the-top it almost reads like satire. Is KDE really that great? (Using Gnome under Ubuntu - no complaints here. But I also am not sure what KDE is giving you. Control over look-and-feel of the windowing environment? Default utility applications? Perhaps a desktop API thick-client programmers can write against?
KDE is really that great. Used a GNOME desktop on Fedora recently and there were so many simple features missing.
Windowing rules for one. Simple example: Firefox picture-in-picture. On KDE I have a windowing rule so that if from any firefox window playing video I hit the picture-in-picture button the picture-in-picture window becomes a certain size, goes to a certain placement on my monitors, stays on top of all windows, and is visible across all virtual desktops.
Ability to control the layout of my virtual desktops is also incredibly useful to me. (I use a 3x3 grid, so switching from my "main" task in the center to any one of 4 sub-tasks up-down-left-right is easy, and my universal tasks (chat/email/etc) go in the 4 corners.
KDE puts you in control, and gives you a LOT of control. IMO GNOME feels much more windows/mac in it's design philosophy. "We know best, do it the way we let you."
And as far as an API, yes, also that. With the Plasma desktop plugin Firefox remembers which virtual desktop each window is supposed to go to, so I have no issue rebooting with 5 or 6 different windows open.
I apologize if it sounds over the top, but for my use-cases at least the level of control and "just does the right thing" really do stand out above the alternatives.
+1 to this. If Signal drops Android SMS support, I suspect it'll create friction within my friend group that uses it. I do not want yet another app for just text messages. No thank you.
No. Secrecy will have backdoor keys, but that is not what walled garden means: it means more like people have no power over the decisions made in the garden.
"... those designated as gatekeepers will have to:
allow third parties to inter-operate with their own services, meaning that smaller platforms will be able to request that dominant messaging platforms enable their users to exchange messages, send voice messages or files across messaging apps. ..."
Thank you, I should have clarified that in the original comment.
(I admit that I actually edited that original comment, since the link I initially included was more speculative, talking about a law that was expected to pass, and being much more vague about when it might enter into force. The page at the updated link is more definitive, but not as narrowly focused, so I'm glad you managed to isolate the relevant section.)
I think I've burned a lot of social goodwill switching my family and close friends to signal and have no desire to support them through yet another change.
Honestly I'm pretty critical of the Signal app design: from the crypto nonsense, to the removal of chat bubble colors (used to be each person had a color, pretty useful in group chats), to the copious amounts of whitespace that have been linearly increasing for years, to the fact that the design has to change and break familiarity every 6 months or the devs have a stroke.
But I actually like this decision. It makes things less confusing and accidental use of unsecure SMS impossible. The downside is if you still use SMS you have to keep 2 apps, back them up separately, etc.
> "Why do I need 3 apps (Android Messages, Signal, Whatsapp)"
"You need Signal to talk to people on Signal, WhatsApp to talk to people on WhatsApp, and Messages to talk to people on SMS." Seems more straightforward than "use WhatsApp to talk to people on WhatsApp and Signal to talk to people on Signal or SMS; just pay attention to the color of the send button".
This issue for myself and many others is it makes something that used to be transparent, entirely unsupported. The UX is unambiguously worse. I could trust signal to upgrade my texts for me when possible, or not when my contacts were SMS. I don't care about always being encrypted 100% of the time. Signal was that perfect tradeoff between privacy, and ease of use, which is exceptionally rare. Providing this tradeoff is what made them popular, them going against it is counterproductive and will hurt them badly. I know this because now I'm considering leaving myself.
I dunno, all my friends use at least a handful of messaging apps (iMessage, FB messenger, Discord, Telegram, SMS). Sure people grumble about a new messaging app but the younger generation seems to not have an issue adopting new things.
I wouldn’t frame it as an issue around adopting new things. Some don’t care, some go with the flow, and some prefer to make active choices about these kinds of things.
I am very intentional and active when it comes to what has push notification privileges. I factor that into my app use consideration. I have multiple email accounts in two different email apps, each that send me notifications. I have Signal, Discord, iMessages and SMS. I have a few Google chat apps. I used to have WhatsApp and Wickr and Telegram. I have Skype, Teams, and two Mattermost servers.
It’s exhausting to constantly switch between these, so over the course of a few years I’ve been very clear in where people can expect to reach me reliably. If you need or want to chat with me on Discord, Skype, or Google whatever you need to send me an iMessage, SMS, Mattermost, or Signal message. Sending me a message anywhere else will get you a response only the next time I open that app. That only happens when someone specifically asks.
I’m OK with having 63847394038 chat and video calling apps, but I’m not OK with being instantaneously notified by an infinity such apps. I can’t be that available.
Yeah, I can understand that; but I've brought over various older family members, and non-tech friends (as in people that wouldn't have ever heard the words Discord or Telegram before in their lives) to Signal. That's who this will impact most.
And while older generations might be less willing to use a high number of apps side by side, having one kind of message in one app and the other kind of message in the other app is still much less confusing to them than dealing with the subtleties of multiprotocol if everything is forced through the single one-size-fits-all interface of a messenger that tries to do SMS on the side.
The US seems to be the only place where everyone uses iMessage, so Android users have to use SMS and suffer the bizarre shaming of the green bubble. In most countries outside the US, WhatsApp seems to be the default. SMS is just legacy 2FA messages, and various other transactional messages like parcel delivery notifications.
I'm guessing. Though I'm in the US but also in grad school. With a large number of foreign students there are similarly a large number of WhatsApp, Telegram, and WeChat users. I suspect this problem is very Americentric. India was able to get all its old people to use WhatsApp and multiple apps, I think there is a bit of an overreaction going on here. You'd think the world is ending for a feature most people didn't know existed (despite it being a prompt during signup).
Is the snark and reducing of the parent comment to a “bla bla bla” really necessary? We’re all adults here, we can make a point while treating the other person with respect.
I'm very upset by this decision. I've been using Signal as my SMS app for a very long time.
Messages that I would have sent via SMS currently will automatically get sent via Signal if the person I'm sending to has started using Signal without my knowledge. This has happened in several instances where I was pleasantly surprised to see a friend had started using Signal. Now that I'm forced into a separate SMS app, this will no longer be a possibility. I certainly won't be firing up Signal to see if a contact has joined before sending them an SMS.
This. Now you have to remember who is in Signal and who isn't. All because apparently the double-check mark for messages between Signal users and the unlocked icon for SMS messages is too hard to comprehend. SMH.
If I understand this, if I use SMS, I can send to everyone. If I use Signal, I can send to Signal users only. But I don't remember who's on Signal, and who's not. So I guess I will stop using Signal.
If I want to message someone I open the contact and click on one of the messengers that are listed for the phone number. Why would I leave the memorizing to my brain?
Huh. I've never used contacts that way. I suppose it could work but that's a new extra step. My Contacts list is gigantic and full of bullshit I don't care about because it's sync'd from work and flooded with people I don't know. Usually I just find the conversation from the chronological list (which is more of how I remember things). Maybe there's some way to sort contacts by recent use? It just seems like that's leaking metadata to push all of that context into Contacts. Anyway that seems maybe plausible if it can index or springboard to convos in other apps.
Theoretically couldn't an Android app be built using notification access to track notification history and coalate messaging notifications (combining notifications from all/selected/configured) messaging apps? That sort of a "messaging hub" could be even better, frankly.
What'd be even better is if there was a central "messaging" interface where all the various implementations can register as messaging service providers and all your conversations end up in one place.
Off topic slightly, but it amazes me how much SMS is used in outside my country (maybe just US?). I literally never SMS any personal contacts, usually WhatsApp. Even business stuff, sometimes initial contact may be SMS and then could often move to WhatsApp.
I use signal with a small circle of friends, but no one I know uses SMS anymore.
I have WhatsApp installed for two different group chats, Google Chat for another couple of group chats, but apart from that SMS is the standard here (which means opportunistically iMessage / Signal). I'm in Australia, where SMS typically have no per-message cost (the only thing that's charged per use on most mobile plans here is data and international calls).
In Ireland, an SMS used to cost 13c, if I remember correctly.
So when apps like viber, WhatsApp, etc came on the scene, people jumped on them quickly and completely stopped texting.
This was before they even had voice calling.
With voice calling, it is also popular, as you can call someone irrespective of what country they are in and not worry about roaming or international call charges (even though in the EU now we do not have roaming charges anymore).
Some people even use WhatsApp for normal calls over normal cellular calls!
I have been receiving notifications that a person in my contact list is now using Signal for years.
Apart from that, your use case has another possible issue.
If a person stops using Signal, your messages will go to the void until Signal actually removes the user and your client switches back to SMS. This has caused a lot of confusion for some of my friends when I switched my signal account to a different phone number.
I think it's more reliable to use Signal for Signal.
Well this is also a problem. As it's said in the article, you risk getting charged for an SMS, that in some countries are expensive, most mobile plan in my country have 30+Gb for 7 euros at month, but SMS are 20 cent *EACH*. Practically in my country nobody uses SMS, and SMS are used only to receive 2 factor authentication codes (and spam).
Anyway a normal person already uses multiple messaging applications: WhatsApp, Telegram, Facebook Messanger, Instagram direct messages, the good old email, SMS (I guess somebody they are still used reading the comments), adding Signal it's not that big deal.
... and there's the reason I will likely stop using Signal?
Signal was always one of those "win-win" apps, get more security when it's available and I don't have to worry about adding to the giant bucket of messaging apps.
They were a paragon of putting the user first and I was a strong supporter... but now... Why not Telegram? Or anything else?
I don't need the security, it was nice-to-have. Having to switch between Signal and other apps is a heavy amount of friction.
> ... and there's the reason I will likely stop using Signal?
> Signal was always one of those "win-win" apps, get more security when it's available and I don't have to worry about adding to the giant bucket of messaging apps.
Same here. I see no reason to continue using Signal if they do this.
I agree. I picked signal over deltachat to replace group MMS threads because it was less startup friction than getting everyone to login to their email accounts on a mobile account since they got SMSes for free.
Now? Delta chat is looking plenty fine for doing private group chats.
My threat model is not nation states watching my metadata, I have horrible opsec for that. My threat model is discord and whatsapp etc. tossing me and my chat groups off a cliff at their sole discretion.
Signal gave me control over chat groups, and integrated with SMS as a bonus. Now? If I'm gonna have to deal with a separate SMS app anyways, I might as well use delta chat where I know my messages are automatically backed up in my email account.
> Telegram is absolutely the worst when it comes to privacy
Really? Telegram never said that they don't store your messages on cloud, they said that they do not sell your data or share it with third parties for profit.
Telegram has received a very good score on PrivacySpy (https://privacyspy.org), in fact better than any other messaging app. Telegram is good from a regular privacy perspective unless your threat model involves fearing cloud convenience.
If you're someone who requires spy-level opsec, you should be using Threema, Session or Speek. Maybe even a self-hosted XMPP instance.
Telegram is good at what it does and it states it very clearly. It does not lie about the things it does and it is open source. All while not selling user data, not manipulating user behavior through algorithms or censoring media by calculating hashes and providing what's arguably the most feature rich messaging app on the planet for free with a verifiable source code.
Also, be careful with what you're suggesting. Not only have Matrix servers been hacked twice but matrix also leaks metadata. If you're seriously suggesting true anonymity (not consenting privacy) then Matrix is not a good option.
> pp. Telegram is good from a regular privacy perspective unless your threat model involves fearing cloud convenience.
Telegram stores almost everything online without E2EE.
> Not only have Matrix servers been hacked twice but matrix also leaks metadata.
Even Signal leaks meta data.
> If you're seriously suggesting true anonymity (not consenting privacy) then Matrix is not a good option.
Out of Matrix, Telegram and Signal, Matrix is the best option. It is the only one not making you share your phone number giving you anonymity up to your IP address.
and yet I just did. Can we please stop confusing privacy and anonymity?
Your claims about Telegram being bad for privacy are baseless. Your concerns about messages is valid but it in no way compromises privacy because:
1. No telegram employee can read any messages. They use distributed key generation to encrypt data on servers which means no single server has access to decryption keys and all the servers are in different jurisdictions.
2. They do not sell message content data. If you can prove it, you can go ahead with a lawsuit and win a hefty sum.
3. They do not compromise security. They do not use E2EE by default. Their threat model and vision for a messaging platform is different than yours.
4. Telegram has never given message content for a court order. As mentioned in the privacy policy, they give out only the phone number and IP Address only in case of terrorism or child abuse and only when there's a court order from a country of a higher democratic index.
5. If you truly believe Telegram is bad for privacy even after all the evidence from FBI itself and PrivacySpy giving it a higher score than Signal, then please go ahead and sue them because surely they can't have a good privacy policy and bad privacy at the same time.
> 1. No telegram employee can read any messages. They use distributed key generation to encrypt data on servers which means no single server has access to decryption keys and all the servers are in different jurisdictions.
This is wrong. First, reported messages (via id) are read by employees. Second, regardless of your claims, Telegram can easily write a service which has access to plain text messages.
> 2. They do not sell message content data. If you can prove it, you can go ahead with a lawsuit and win a hefty sum.
How about you prove your claims? I hardly can bring them to justice when even the police doesn't have immediate access to them.
> 3. They do not compromise security. They do not use E2EE by default. Their threat model and vision for a messaging platform is different than yours.
Actually, they do by not using E2EE by default and providing bad encryption possiblities.
> 4. Telegram has never given message content for a court order. As mentioned in the privacy policy, they give out only the phone number and IP Address only in case of terrorism or child abuse and only when there's a court order from a country of a higher democratic index.
no idea about that
> 5. If you truly believe Telegram is bad for privacy even after all the evidence from FBI itself and PrivacySpy giving it a higher score than Signal, then please go ahead and sue them because surely they can't have a good privacy policy and bad privacy at the same time.
Since when is "bad for privacy" a reason for suing? The quality of a privacy policy doesn't have anything to do with privacy itself btw.
> and yet I just did. Can we please stop confusing privacy and anonymity?
> Actually, they do by not using E2EE by default and providing bad encryption possiblities.
These are again baseless claims. If you think MTProto 2.0, an encryption algorithm that has been audited multiple times by independent researchers is 'bad encryption', I'd like for you to prove it. Obviously, if you can prove it's bad, you could let Telegram know and win a bounty.
> How about you prove your claims? I hardly can bring them to justice when even the police doesn't have immediate access to them.
The burden of proof is not me as I did not make any claims, I simply restated what's on the Telegram website.
Even the FBI, Iran or Russian government couldn't bribe them so I do trust Telegram to not backdown on their statement and philosophy about not selling or using userdata for profit. https://twitter.com/durov/status/912812889236475904
> Second, regardless of your claims, Telegram can easily write a service which has access to plain text messages.
You do know even Signal could add a keylogger service to read message content right? I don't suppose their Google Play Store version has reproducible builds. See how easily arguments like these break down? You can almost assume anything and claim almost anything. As I said, these are baseless claims and assumptions. I'm only interested in the objective truth at the moment, not assumptions or guesses.
> Since when is "bad for privacy" a reason for suing?
You're suggesting Telegram's privacy policy is in direct violation of their privacy practices which is illegal. This is a huge claim, if you can prove it you should sue them, I'd honestly do that if I were you.
> I didn't, did I? please explain
Privacy is about choosing what to share, not about sharing nothing. You seem to lie more on the anonymity side of the argument than privacy rights. You're fighting for anonymity, not privacy if you claim malicious intent on Telegram's part because as I showed earlier, their privacy practices and security are totally A OK.
In reality almost no one bothers with secret chats (no syncing between devices, no backup and no group chat possible). Instead everything is stored online without E2E encryption, i.e. perfectly readable for the service provider.
I see, that's the 1 on 1 chats that are explicitly configured as secret. So by default for 1 on 1 chats and for all group chats the keys are stored on the server.
Correct.
But in the case of matrix you can host them in your home if you want, or maybe on your phone(they are still checking if this is possible or not)
Maybe I am capable to do so (although I already host an XMPP server, so Matrix is rather redudant) but expecting everyone to self-host is obviously not realistic.
Currently Matrix is operating in a way that larger instances aggregate private messages from bridges in plain text. Those messages would have stayed encrypted and secure if people didn't use Matrix.
> Currently Matrix is operating in a way that larger instances aggregate private messages from bridges in plain text.
That's not true in general. For most if not all messengers (at least encrypted ones), there is the option to use Bridge-to-End encryption on the Matrix side, which doesn't give the homeserver any possibility to inspect messages let alone aggregate them.
If I understand correctly, the Matrix encryption and message format is not compatible to other protocols so all messages must be decrypted and converted on the homeserver which runs the bridge.
The bridge and the home server are separate things.
So if you run the home server somewhere untrusted, but the bridge in your home with E2EE enabled, the decryption AFAIK happens in the bridge, and the home server doesn't see anything.
I doubt many users will host their own bridge. Also at least for XMPP I was not able to encrypt messages to Matrix users end-to-end or bridge-to-end, so I'd still recommend using XMPP with E2EE directly.
"why not anything else" is mostly (for me) because they are a non-profit, and unlikely to be bought by or turn into a megacorp, similar to how wikipedia runs, although they're certainly a mega-something at this point, it still feels a lot less evil than a facebook or a google.
If this were an in-depth announcement with a long and well-structured technical justification attached, I could understand. Though I suspect I'd likely disagree with the decision, I could probably accept it as a simple different of opinion if the arguments were evidently well-thought-through and considered.
This blog-post is so lightweight. There's no technical analysis. There's barely any justification. Yes we know SMS is insecure and yes - it seems plainly obvious that having them in the same UI could pose UX challenges & user confusion issues. So improve the UX and clarify the distinction. Did anyone in Signal consider the userbase or the advantages of this feature at all?
Definitely the end of my Signal usage anyway. It's my main SMS app: my primary motivator is SMS UX, the ability to securely message a tiny subset of my friends is a very nice but ultimately non-vital bonus. Having a separate app for those people isn't worth my while (they're on other platforms I use more).
Yeah, I 100% agree with you about that and the more I've been digesting that explanation, the more I see why this is the correct thing to do.
Hopefully this even frees them to do things we've wanted for a long time... like not being tied to a phone number and offering better features than RCS/iMessage. Maybe even having multiple independent profiles/pseudonyms for compartmentalization.
That's how Signal could be growing the base and interop with SMS/MMS/RCS cruft on one platform will always lack the killer feature and be irrelevant to the other platforms. If Signal were better than SMS/RCS/iMessage people will just use it for those reasons in addition to the security and privacy.
And having just installed the beta and used the SMS export and allowed it to purge all of SMS content from Signal into Google Messages it actually sort of is nice that the app is now ONLY the "Signal" context. I'm... actually pretty okay with SMS belonging to the "Stuff that Creepy Companies Like Google Know" context.
Basically this just does what Signal already does in iOS: it must compete with the native messaging client. Google is already playing RCS as SMS upgrade and Signal is making the correct strategic decision to not make a play for RCS. SMS support is just going to lead to whining about lack of RCS. The bottom line is both Apple and Google are out to kill SMS. With SMS gone, Signal can just move on to feature parity with iMessage and beyond while leapfrogging whatever messaging clusterfuck Google keeps producing. Google can have SMS for all I care. We can't have iMessage on Android, but we can have Signal on both Android and iOS.
They definetly should publish those points out in the open. After reading this, it just make sense they are dropping SMS, as infuriating as it is. Thanks for the link.
I guess to be fair it lets them design and support a single UX since iOS doesn't allow them to have SMS in the UX. That could have been a good argument.
Of course, they didn't bother make that argument.
And in the SMS domain Google Messages really does get annoying with the whole Google Messages vs iMessage and how nothing Google is doing with RCS benefits anyone except Google. As Google continues its war on SMS and force migration of everyone to RCS, Signal users on Android end up being the red-headed step child. That also is a good technical/strategic argument for ditching SMS.
But, again, not one that they even bothered make.
And there's always been the "tied to a phone number" issue that's been the #1 complaint about Signal. And once untethered from SMS who cares about phone numbers anymore.
Once again, not even a case they bothered to make.
"The answer is this: They dont want to add RCS support or spend the time to do it."
... confirmed by Signal in their discussion thread:
"... and Signal can’t add RCS support because there’s no RCS API on Android. Honestly, the days of any third-party SMS app are numbered."
I guess I misunderstood RCS. I thought the whole point of RCS was to be used on Android and to allow disparate third parties to use it as an open standard.
Where is the RCS API if not on Android ? Who is supposed to use RCS ?
Google also restricts their specific flavor of RCS (or at least they did awhile ago). I wanted to keep using Textra SMS but they never let Textra into RCS land.
Improving UI/UX around to clarify the SMS function is insecure is almost impossible. Google did research around SSL cert warnings a few years back, their conclusion was that people don’t read and just dismiss warnings, no mater what UI was. A frightening percentage of people also think the security padlock icon is actually a handbag.
Most people simply lack the technical basis to understand the security implications of sms. And for Signal to be a secure messaging system by default SMS needs to be removed.
That's assuming a lot of context. Your talking about a tiny icon next to the address bar in a browser. Of course people didn't always know what that was!
Signal's primary feature is encrypted messaging. You don't get it without at least seeing the word "encrypted" somewhere.
And that doesn't get clarified by UI that distinguishes between encrypted messages and SMS, because Telegram doesn't have such a thing to distinguish between.
My point is that all of this is orthogonal to whether Signal can successfully make UI show users when they are sending encrypted messages vs unencrypted SMS.
Most of the confusion you are citing is about whether an app does encryption or not, and that is a totally distinct problem domain.
The only similarity between these two UX scenarios is that they involve encrypted network protocols. From a user standpoint there's no similarities.
Firstly, the messaging decision is presented to the user before an action (send SMS/Signal). It's capable of blocking and takes place as part of an active use flow where the user is trying to complete a task. With browsers, the differentiation in UI is displayed after a user action. It doesn't block and the user doesn't require interaction to achieve any goal. Why on earth should they pay any attention to it?
Secondly, the UX for messaging is an equivalent paths binary decision: you're asking people to choose A or B. There isn't an inherent default so a user doesn't start out with a bias toward one or the other. They can easily be required to read to proceed.
With browsers it's a yes/no binary decision: the default (yes) is insecure (for an insecure website). It requires no action from the user. The secure option (no, leave) asks the user to do something. It's a choice between inaction (insecure) or action (secure). That's heavily stacked.
Lastly, even the context surrounding the apps themselves is incomparably different. One is a security upgrade of an application everyone's been using for decades (often unknowingly; "the icon for the internet"). The other is an app people consciously download and install explicitly for security reasons (regardless of whether they understand those security reasons it's at least the motivating factor).
The people you talk about see no sense to use signal at all. So why should they install it when they have SMS? And when Signal is installed, why should the change the app and use signal instead of SMS?
What kind of technical analysis would you be looking for? Reading the post, it seems like their analysis came down to (1) fundamental values, i.e. not including insecure communications within an app when they've built their brand around being secure, and (2) UX confusion resulting in additional SMS costs and/or inadvertent data leakage. The former is a straightforward question of product strategy. Are you looking for e.g. some numbers from their UX research? This doesn't seem to ultimately be a decision about underlying technology.
> Definitely the end of my Signal usage anyway. It's my main SMS app: my primary motivator is SMS UX, the ability to securely message a tiny subset of my friends is a very nice but ultimately non-vital bonus.
I think this is the crux of it. Your primary motivator may be for a better SMS UX. But Signal's primary motivator is to provide universal secure messaging, but your typical use of Signal doesn't do that. So it's no surprise that their plans mismatch your expectations.
All centralised & protocol-locked messaging apps are subject to network effect. People moving away from Signal doesn't help the goal of universal secure messaging, regardless of whether those people are you or I.
That said, it seems they're between a rock & a hard place here since Google are defacto deprecating support for 3rd-party SMS apps.
That's just wishful thinking. Any opportunity for a network effect to assist with Signal's adoption is long gone. It never managed to hit the necessary threshold and without platform ownership, there's little chance it will.
> The most important reason for us to remove SMS support from Android is that plaintext SMS messages are inherently insecure.
This is an incredibly bad reason to remove SMS support. Sure, the fact "plaintext SMS messages are inherently insecure" is true, but the implication is not "remove SMS support".
Most people are motivated strongly by convenience. Signal is convenient because of its use as a drop-in replacement for your existing SMS client, so people use it, which increases their personal privacy and security. Removing SMS support will directly and substantially reduce Signal usage, and therefore both of those things.
The solution to "SMS is insecure" is pretty obviously "make a warning message telling users that", which also solves their second problem:
> This brings us to our second reason: we’ve heard repeatedly from people who’ve been hit with high messaging fees after assuming that the SMS messages they were sending were Signal messages, only to find out that they were using SMS, and being charged by their telecom provider.
...and the third problem:
> Third, there are serious UX and design implications to inviting SMS messages to live beside Signal messages in the Signal interface.
This is ridiculous. You're not making a paid product where if your app doesn't look perfect people won't use it - you're making a messaging app, and slightly ugly workarounds are perfectly OK.
> It’s important that people don’t mistake SMS messages sent or received via the Signal interface as secure and private when in fact they are not.
THEN DESIGN THE APP THAT WAY. IT'S NOT THAT HARD.
This post is a travesty, and the reasoning contained inside is completely insane.
Wikipedia says that Moxie is still on the Signal Board of Directors, but I find it hard to believe that he would let something this crazy go through.
> Wikipedia says that Moxie is still on the Signal Board of Directors, but I find it hard to believe that he would let something this crazy go through.
IIRC I read (some years ago) that Moxie wasn't really convinced that SMS support should stay in Signal-Android, either.
He was definitely against the encryption-over-SMS feature of TextSecure as Android and smartphones more broadly grew in marketshare. He also wrote the blog post on how it doesn't matter if you have multiple messaging apps (or federation between them) because the notification area of your phone is the modern federation engine. I may be paraphrasing a bit heavily but the post is at [1].
I agree I can see him being at least OK with removing SMS but it seems at odds with what I felt was his overall view of "get the most people the most security we can" and by extension increasing the number of people using secure messaging services to normalize it so simply using encryption isn't seen as an outlier. The latter part is closer to moot now more than ever before with WhatsApp being E2E by default and Apple having huge marketshare in some markets with iMessage.
The ecosystem has kept moving, and Signal has not been keeping pace.
Whether Signal can catch up to its open and closed source competition remains an open question, and I don't think features like Signal Stories are going to be what helps them start nipping at the heels of the other much larger competition.
> I agree I can see him being at least OK with removing SMS but it seems at odds with what I felt was his overall view of "get the most people the most security we can"
They already tried putting a small light grey "unlocked" icon on messages. If that doesn't scream "SMS", nothing does. All available options exhausted. Time to throw in the towel.
This was one of the core features of using Signal for me. I wish they had implemented RCS and more features for SMS instead of removing it. I'm very disappointed with this feature.
As a side note, I'm on the beta, and recently got "Signal Stories". This immensely annoyed me, and had to dig through to remove it (since it wasn't obvious). After the whole crypto thing and these decisions, it might be time to find another secure messaging app.
Because it's a carrier-owned upsell feature. It's nuts that people even consider a non-Internet protocol that applications can't implement, and users have to get permission to use — from the same carriers that charge for bytes of crappy size-limited MMS messages in the same ballpark as delivery of physical letters.
SMS is carrier owned and had an API. What you're saying doesn't surprise me, but I guess an RCS client written by Google and can't talk to iPhones isn't exactly an upsell.
Yeah, I don't know what the Google/Jibe/carrier relationships are, but Google made an exception for Samsung[1], which I suppose could've been carrier approved too. Looking more at it, it seems like Google Messages is still the RCS client, and it's exposing messages to Samsung's app.
Honestly, it's kind of hard to blame Signal and/or Google if carriers are involved here. I mostly use Signal and Element, so if those two had a single client, that'd be fine. But, I think Signal probably still favors a centralized system, and they have added ongoing call transfers from and to mobile/desktop, and now Stories and the money transfer (not a fan), so I suppose it's up to the users to choose.
It's already bad enough that I would never be able to convince family today to switch to Signal due to the removal of SMS history importing and now you want to remove the ability to send/receive SMS via Signal too? Good job guaranteeing you just cratered any additional growth of your userbase.
I've always wondered how companies become so blind to what their userbase actually wants and needs (looking at the majority of the rest of the comments here that seem to echo my sentiment as well) that we end up in situations like this. I guess "you die a hero or live long enough to become the villain" applies to apps too.
Literally the only reason I recommend others and use Signal myself?
Seriously, Signal doesn't have the userbase to drop SMS support. All my Signal contacts use WhatsApp or Telegram that I already have installed. I use signal mostly as a SMS app, secondly as E2E communication. It will be easier to uninstall Signal.
Seriously. I don't want to use another locked in messenger app, that everyone else must use or I won't get their messages. I use signal because it's secure, but also because it's low friction and seamless into SMS if the other user doesn't have signal. This is another step in the wrong direction for Signal.
The complete lack of awareness of this decision is astounding, the userbase is about to disappear. Those of us that convinced non-technical friends and family to use Signal are now expected to explain to them how to juggle 2 messaging apps? yeah it's not happening, the uninstall rate is going to be huge and there will be no recovery of those users
> If you do use Signal as your default SMS app on Android, you will need to select a new default SMS app on your phone. If you want to keep them, you’ll also need to export your SMS messages from Signal into that new app.
This messaging seems a little tone-deaf, given that there is no way to export SMS messages from Signal. Apparently it's possible, using a third-party piece of software, to decrypt your backups and extract the messages, but that's not exactly a reasonable thing to expect people to do.
One of the reasons I liked Signal was because it was easy to get normal people to start using it, because they could just set it up as their SMS app, and continue life as normal, just getting the benefits of encryption for any of their contacts that were also using Signal. Now there's not notably any reason to use Signal as opposed to, say, Matrix.
Confirmed. Turned on the beta program and exported ~1000 text messages over to Google Messages. Settings->Chats->SMS and MMS->Export. Involves changing the default app for SMS between Signal and Google Messages.
I can't fathom the cognitive dissonance of Signal community leaders actually suggesting their Android users to migrate to Google Messages. We're going from a best-in-class privacy app to an app best known for it's vendor's data hoarding. All because they think we should have the same sub-par SMS UX that iOS users are subjected to...
I completely disagree and am disappointed in this decision. One app on my phone to handle all my messages is easier than making a context switch per-contact.
I also think it'll hurt the value proposition when getting people to join signal. Not overcomplicating the messaging scenario was a big winner to do that.
> all my messages is easier than making a context switch per-contact
A user already has:
- WhatsApp
- Telegram
- Facebook Messanger
- Instagram that has direct messages
- the good old email, or better, many of them
- Microsoft Teams for company communications
- Discord for communications with group of friends
- the old SMS (that I didn't even know that in some parts of the world were still used, since I receive them only for 2 factor codes, notifications about my card transactions, and spam)
Adding another app is that a big deal? By the way I don't use Signal, but not for the reason of not having another app on the phone, just because I don't know anyone that has it and actively use it.
You're right that there are many messengers available and Signal will, at best, be one of many that people juggle based on who they are contacting.
That's why it was a huge advantage that, on Android, Signal could replace a SMS client. You weren't adding _yet another_ messenger to the list, you were replacing the SMS client with one that could send secure messages. That made "switching" to Signal (which, ofc, was not a switch at all for my friends who use SMS) much easier for me. I could continue texting my friends and seamlessly switch to secure messaging if they ever got signal.
Contrast this with my friends who kept their old SMS client who reliably forget to check / use signal and generally tend to go back to texting me in a few weeks. Even if you send 0 signal messages for a long time, by switching you SMS client you are already setup to receive them and will habitually open an app that supports E2E encryption.
For example - Facebook Messenger also supports sending and receiving SMS messages - likely because they've done the research and found it drives adoption.
Does anyone actually have all those? I certainly don't. I have Signal, Element, Telegram and I even think that's excessive. I can at least manage it, most RL contacts I know would not.
Well, not all of them, but usually you expect a young person to have all of them. Except Microsoft Teams, the company may use other media, and Discord.
Hell, even my aunt that doesn't know nothing about technology has WhatsApp and had me install Telegram because the church opened a channel on it!
If you don't use Signal do you really have any skin in the game anyway? People obviously care that it has SMS capabilities and it is widely used for that, so naturally people will be upset.
This does not represent all users. My family has boomers who literally use just Signal (for 80% SMS) and FB Messenger. For many people, less apps is an important feature.
I do not like this decision. Using Signal as a main SMS provider makes it easier for me to collect all of my messages in one place. Now I have to, YET AGAIN, download an SMS app for use while keeping Signal active.
I'm glad privacy is becoming mainstream but dislike lowering the bar for adoption to where it profoundly affects users.
It makes me yearn for the days with Pidgin where I had IRC, Google Chat (XMPP back then), AOL and whatever else chat protocols all running through the same client.
That's what is nice about signals implementation is it stands. It supports acting as the SMS default app on android and defaults to signal when it can.
> That's what is nice about signals implementation is it stands.
Sure they handle SMS, but the real problem here is that Signal is just another walled garden: they have an overtly negative stance towards alternative clients, while also having very bad support for anything besides android/ios: they have a bad desktop client and they don't have a nice library. Altogether this means that Signal is overtly and willingly against things like Pidgin / multi-protocol clients or overlay, which is what the users want (ie not caring about protocols).
Signal doesn't want to deal with SMS anymore, which from an engineering and high-stakes security pov is a completely valid decision. Yet if it had clean and open local API or a simple and portable client library, or had a stable server API, then someone else could provide multi-protocol clients, tailored to each platform in a secure and stable way.
One thing I liked about those multi-protocol clients is that some of them supported the OTR libraries for E2EE encrypted messages regardless of platform used. A couple of the implementations would automatically handshake with others to see if they supported OTR.
FYI Part of your website is broken on Firefox for Android. (Broken layout, content not shown etc.)
Now to my actual question: How is Beeper compatible with the ToS of platforms like Instagram and Facebook that, to my knowledge, don't allow their users to use 3rd-party apps? Case in point: I recently wanted to use a FOSS 3rd-party messaging app for Instagram and my account got promptly banned.
Question 2: Do you support full message backups in a well-documented format?
Unfortunately, at least in the U.S., most inter-OS text-messaging is still done via SMS. Signal was godsend in this field because I can slowly convince my network to switch to Signal (and this in turn had a recursive network effect as then they would do similarly). This change will mean Signal will become another bucket on my phone (along with WhatsApp) where I can talk to only a select few of my contacts.
This feels like a slap in the face. I get the privacy ramifications, but one of the really strong aspects of Signal to me was to go all-in on privacy when needed, and default to something sensible when it wasn't. I'll definitely need to reconsider whether or not to continue my monthly donation, and I don't like that at all.
This seems to be a "bug or feature" situation where the answer depends on the user profile. The ability for messages to leave the Signal app in plaintext SMS is a "feature" for users whose top needs include a single-app UX, and a "bug" for users whose top needs include an app that is foolproof E2EE (so users don't have to consciously pay attention to which conversations are Signal-native vs SMS). Maybe SMS support could be an opt-in feature, to accommodate both groups?
From my perspective (and I am NOT speaking for anybody else) this is an improvement. I already have multiple messaging apps installed, and when I click send on a Signal message I expect it to go end-to-end encrypted or not go at all. But I am not the only user profile.
I agree with your perspective - for me Signal is yet another (more secure) internet messaging app on my phone, and I'm happy that way. I wouldn't want it to have anything to do with SMS, no more than I want FB messenger to start handling my SMS (which it does offer at installation time). Plus, having used Signal since the TextSecure days, I saw the SMS feature the same way the announcement seems to characterize it: as old tech debt waiting to get dropped. After all, I don't think Signal for iOS ever had SMS ability.
And to your main point, I hadn't even considered before seeing this comment thread that anybody felt differently, let alone so strongly. Really illustrates how differently people think about the same app.
Never used signal, so I don't know it's UI language. Couldn't they just put a red unlocked padlock symbol next to the send button, if the sent message is going to be over an insecure channel (SMS in this case)? Maybe they already had something like this, so sorry for my ignorance.
The SMS symbol was a dull grey circle with a paper plane and an open lock, while the encrypted message was a big blue circle with a lock in the center. You had to choose purposefully with a long press on that button to use SMS if that is what you wanted, and the recipient was using signal.
UX elements already exist to tell the user whether they're encrypted or not, removing the ability to send unencrypted shrinks their userbase while ensuring the apps adoption plummets.
No, and thank you for asking. No one addressed what to me is the main point, which is that Signal is handling complaints from people who were being charged for SMS and didn't know that would happen. People these days are often not at all civil when dealing with support issues like this, and for that reason alone I could imagine I'd drop SMS.
> We have now reached the point where SMS support no longer makes sense.
What a laughable, out of touch suggestion. Did anyone at Signal actually ask the community what they thought about removing SMS support?
Seriously, this decision is going to kill Signal app. It will halt the majority of growth as evangelists such as myself can no longer recommend it with a straight face. Signal is supposed to enhance the messaging experience, not replace it.
I think Signal thinks they can take on the WhatsApp market, completely misunderstanding why that market didn't choose Signal in the first place. The products serve two completely different user needs, and are highly geographically segregated.
What the heck is going on over at Signal Foundation?
Personally I think supporting SMS is what's "out of touch". It's a legacy system that a huge chunk of their users won't have ever used in the app except possibly accidentally.
It's no more legacy than FM Radio. Been around since the 30s, is undeniably ancient, but it's not legacy. SMS is not going anywhere in many parts of the world, it's actively supported and actively used.
> We have now reached the point where SMS support no longer makes sense
That is hard to swallow, being able to quickly send a message through SMS to the same receiver in emergency situations* was quite handy.
*like when you're at a protest and the tower is overloaded, or you're on a remote location and you see that the Signal message doesn't get through because of lack of 3G/LTE connectivity.
Just a guess, but this likely has something to do with 10DLC and/or Toll Free Verification and all of the complexities that are being pushed by the carriers for users to register their numbers and even pay to use if you want to use 10DLC.
I believe "SMS support" just means Signal can act as your SMS client using your existing modem & SIM card (something possible on Android), so from the carrier and phone network perspective there is no difference between this and using the stock SMS app.
This change will have fewer people use Signal. One reason I was able to convince friends and family to start using it is because it is so seamless. I fear that with this change, Signal for most users will simply become unused, resulting in less e2e encrypted messaging overall.
How do you know if your contacts use Signal and know to use that app instead of SMS/Messages or whatever?
With the SMS integration it was pretty easy because it would just switch over if the other person had Signal or if/when they signed up in the future.
What's the workflow now? Manually ask them on SMS if they use Signal? Just try it and see if it works?
This sounds like one of those "Don't Worry! Rejoice! We're breaking your things!" announcements that hasn't even thought about how people use Signal IRL.
I'm going to stop my monthly subscription to Signal Foundation.
I have instead come around to support this move 200% and have instead doubled my monthly subscription. The explanation at the blog post is an abomination, however.
Leave SMS and all its shitty successors for Apple and Google and carriers to kill/maintain.
The blog post needs to be shelved and redone as every listed reason feels post hoc while the reasons listed at that link ([1] for anyone who dislikes friction) are grounded in reality and show Signal being proactive.
What about the issue it mentions of replies to SMS coming over RCS and never arriving in SMS apps? I'm inclined to agree with the assessment that all 3rd-party SMS on Android are toast. Eventually only the device's/carrier's stock messanger will be reliable.
I think the real rationale for this change is signal believes this will push user adoption.
If User A (who uses the signal app) regularly communicates with User B (who doesn't), then this change might encourage User A to ask User B to join signal. It makes a stronger network effect, and will increase viral growth.
However, I think the Signal team is misguided, and in fact they will just lose users who don't want one more app to manage.
> If User A (who uses the signal app) regularly communicates with User B (who doesn't), then this change might encourage User A to ask User B to join signal. It makes a stronger network effect, and will increase viral growth.
Conversely, the inconvenience of having multiple messaging apps could cause User A to stop using Signal. Look at what happened with Hangouts when they dropped SMS support.
SMS support is literally how I got my family to switch to Signal in the first place. None of the non-techies want to switch apps or have to send the same message out multiple times in order to reach their friends and family. Having an app that provides privacy when able and still works for those not yet onboard was a godsend.
> So I guess the TL;DR is: SMS is on it’s way out in general, and in a world where Signal supports SMS, all of SMS’s shortcomings are often attributed to Signal itself, all while confusing people into thinking their SMS’s are secure.
I was responsible for roughly a dozen Signal converts. They will all be uninstalling when this takes effect. It seems like everyone else in this thread is saying the same.
Prepare for a mass exodus of users.
None of us want to be responsible for training/tech support for how to use 2 messaging apps for non-technical users
What dicks. I'm not looking forward to playing tech support for all the non-technical people I convinced to use Signal. Thanks for confirming everyone's suspicions about my weird-nerd chat client.
Very disappointing and upsetting. I use Signal as my primary SMS/MMS app on my phone, and use a few Signal chats as well with people. This is going to be really annoying. I'm probably going to just stop using Signal altogether to be honest.
Most people in my social circle use Snapchat or iMessage for "texting", for reference.
One bright side of this is that Android's (Google's) Messages app has been pushing hard on RCS (the intended successor to SMS) and by default now does auto-upgrade to end-to-end encryption with any other messages users. If you're using signal, you don't get that auto-upgrade, so for conversations with anyone using a "default" google phone setup you were actually getting less net security on your comms compared to using the default SMS app.
I noticed this when I got a new phone and hadn't yet enabled signal to handle SMS and opted to stay with it because of how many conversations I had that were auto-E2E, where before they'd just been text messages. I still prefer signal for the people I know use it though. In short you can still use the signal (protocol at least) on messages, so I can understand why signal would do this.
Given that it's Google, are you sure they do true end to end encryption. I would be shocked if they don't have access to the contents of your messages.
They do the same as Whatsapp. Ie. it is proper end to end encryption. Encryption keys can be verified manually. But there is no way to know the app doesn't secretly send the key to the server (although a disassembly of the app could catch them red handed if this were the case).
The big loophole is:
* The messages can be forced to be sent unencrypted if one or other end of the connection doesn't have data connectivity.
* The conversation backups are cleartext, so if either you or the other party has backups enabled, the e2e encryption is kinda pointless.
That sucks. The data fee argument makes no sense – you could just have a setting or warning or something for those who live in places where you have to pay for sms (I know every setting introduces complexity, but I that's got to be nothing compared to the level of engineering needed for all those other fancy features in Signal).
It's completely backwards for me. When I'm out of data for a month, SMSs still work. I've had to press and hold the send button to revert to SMS on many occasions.
What's even worse is that for you (and I) removal of SMS support will mean that out message history will suddenly be inconsistent as existing SMS messages will be removed.
In addition to what everyone else here is saying (this is the most mind-bogglingly stupid idea you could imagine, which will instantly kill the adoption of Signal in the US) I want to point out that the purported reasons for removing this feature would be completely solved by hiding SMS behind a setting. If you want to be EVEN MORE paranoid you could periodically warn users if this setting is enabled, just like they periodically bug you about your pin. The only explanation I can have for this decision is that the real reasons for it have nothing to do with those given.
> After much discussion, we determined that we can no longer continue to invest in accommodating SMS in the Android app while also dedicating the resources we need to make Signal the best messenger out there.
I did not need emoji's, groups, gifs and all the other neat stuff that signal has introduced throughout the years(to varying degrees of success). I had been using it, while none of my friends were. What I did need was a single messenger to handle sms/mms with the default being secure when security was available. I have multiple friends now using it and sadly will revert back to a 100% insecure messenger for my phone for 99% of my messages. The new one will do everything better than signal does except security, so it will have some benefits.
I will be on the lookout for a replacement. I hope signal continues to bring security for entities that need it through the future. I have not looked at tox in a while. I'll check that out again.
> There are three big reasons why we’re removing SMS support for the Android app now: prioritizing security and privacy, ensuring people aren’t hit with unexpected messaging bills, and creating a clear and intelligible user experience for anyone sending messages on Signal.
Pretty weak reasoning to me. Just do what Apple does and color sms messages some other color or whatever. Problem solved.
This is gonna make me drop Signal. I use it as my default sms app and have been very happy with it, but most of my conversations (although most actual messages are Signal) are still over sms so it'll have to go. I can't be bothered to roll a bunch of different apps.
Still, I'm grateful for the work the Signal team has done over the years. Sad to see us part ways!
Hm, not in my app (Android)? Unencrypted messages have an open padlock next to them, whereas encrypted ones have double checkmarks. But no difference in color, even in conversations that contain both.
This kind of thinking is exactly what's killing signal. Why would you want to gatekeep the security of others instead of making t as accessible as possible?
To me this feels like signal not understanding that their intended userbase and their actual userbase are very different, as I can't imagine the number of people that use signal solely for it's e2ee is comparable to the number of people that use it as their sms app.
Well, isn't the point that more people using signal strengthens the privacy/security of everyone? I would be sad if I wanted to send a message to gerty, and I find out he stopped using signal because of this.
It's not like I wouldn't still want to message him, right?
Convenience is extremely powerful in getting the layman to adopt this kind of tech, and I feel like it should be prioritized.
Terrible decision. You don't improve the average person's security posture by increasing the barrier-to-entry of encrypted messaging - and removing SMS support is doing exactly that. Signal -is- was great BECAUSE it made the transition from SMS to Signal so seamless.
It is a US thing. The majority of folks in the US still use SMS for text chat. The alternative is buying into a proprietary platform like WhatsApp, but not everyone you talk to is going to be on WhatsApp/FB Messenger/iMessage, etc, so you end up having 5 different chat apps installed on your phone at any time. SMS has been the only real ubiquitous one despite all its flaws.
I don't live in the US, and SMS support has been essential in convincing people from my parents' generation to give signal a try. I personally also like not ever having to open my phone's stock message app when I once in a while do get an SMS
I don't send many SMSs, but I receive them: package delivery notifications, automatic appointment reminders, a message to say my car has been repaired, 2FA codes.
I found it useful to have all these messages in a single place, although this change probably won't inconvenience me too much. However, I don't see any benefit.
Why the rest of the world has converged on using a free closed source messaging service by a for profit company because of the privacy boggles my mind. I'd rather just stick with SMS if I can't feasibly use signal.
Europe is a big place, but indeed I'd say in most of EU using SMS to communicate between people is a bit old school. I'd say most people have moved to WhatsApp (or equivalent).
SMS is typically used for notifications-style messaging (or spam), so the impact of this change is probably minimal for a lot of people there.
Even chat-bot/support style messaging in the B2C space are moving to WhatsApp (or equivalent).
I use signal as my SMS program and a few people who have signal. if I can't use it as my SMS program I'm not going to keep using it for the handful of people who have signal and will likely just go back to SMS for everyone.
I am unhappy with this change, but I can cope with it. I'm more concerned with my tech-challenged family members who don't understand the distinction between different messaging services or have any understanding of security. Until now, Signal has been good for them because they only need to deal with one application and they get some added security among our group. After this change, I fear they'll just use the SMS app exclusively (out of inertia) and Signal will collect dust.
No doubt they are in a tough spot. Some users will not accept this feature omission. But if what they claim is accurate, the insecure nature of SMS, along with Google's hoarding of their internal RCS APIs makes it tough to be a messaging provider on Android.
I have been using Signal for years. The ability to make it your default SMS client is one of the major drivers of adoption; if someone agrees that privacy matters, and you can point out that the transition to Signal is frictionless and offers all the same features as their existing SMS app, then installing, trying, and liking it become easy. I've brought hundreds of people onto Signal, and being able to give a simple 'yes' to questions about whether it handles SMS is almost always what 'seals the deal.'
Signal is saying that mixing non-secure and secure messages in the same app might cause confusion and security fails, even though the difference is very clearly signalled.
Their argument is bullshit. If users go back to separate messaging apps, chances are those apps will look much the same as Signal (which itself copies the look and feel of the iOS messaging app quite closely). There's a much bigger security risk from users forgetting that they are not in Signal and carelessly pasting & sending information that was supposed to be private or disappear.
Additionally, it creates a bunch of new security risks, allowing third parties who gain possession of a phone to distinguish between conversations that happen over SMS and conversations that happen over Signal, drawing inferences that there is something untoward about the latter.
I cannot understand the constantly changing, er, signals coming from Signal. One month they want to be just like every other messaging app and they're pushing features that hardly anyone has asked for, like sticker packs or crypto payments. Other times they say users are too paranoid for not wanting to expose their phone number/pop up messages about who in the user's address book has installed Signal. Today they're saying that wanting to use Signal for all your messaging needs is somehow anti-privacy.
I find myself wishing it cost money or a small annual subscription so I could vote with my $, because the Signal foundation seems to spend more effort on telling its users that they're wrong than on listening to them.
For anyone that's upset about this, don't focus on the SMS vs. Signal messages UI distinction giving as the primary reason.
That's something a simple UX change could improve and does not justify something so radical. There's probably another reason they are doing this, these are my best guesses:
- Feature parity between Android, iOS and Desktop
- Moving towards Signal accounts tied to usernames instead of phone numbers
- Developer resources (unlikely, since this feature has existed since the beginning and probably requires less maintenance than other features, but I'd pay for a subscription if they kept it)
Whatever the real reason is, make them say so, the reason given is extremely flimsy. If they go through with this it will put lives in danger for people like activists, journalists, and anyone crossing a border who depend on their SMS messages being encrypted at rest (despite SMS not being end-to-end encrypted in transit). Using Signal instead of a different SMS app also prevents other apps from reading your SMS messages. Test this out - switch your default SMS away from Signal and sign into an app requiring SMS verification, and that SMS is probably readable by the app without any interaction on your part.
Cue the Apple-ish response of them saying they're listening while not taking any of their users' valid concerns seriously at all. Very disappointed in them.
Alright, it's fine for a company to remove features, if they are honest about why they are doing so. It's obvious none of the reasons given are due to user complaints. The truth is, they are removing SMS because they don't own the SMS platform (e.g. it's not a walled-garden like WhatsApp). Would it kill them to just be honest? Yes, it's less secure, but no end-user is saying, "Please remove SMS as it's not secure."
I don't think that's true? There are clearly users who are annoyed at being able to send insecure and possibly expensive SMS messages in their secure messaging app.
I started using Signal in 2013, I am afraid that I will be forced to stop using it in 2023. Please change your direction!!! No one wants to use yet another messaging app. Just adding my voice in case some at Signal is reading these.
> supports plain SMS/MMS to function as a unified messenger
So this is now a lie. This decision absolutely goes against how users actually use the software. Tone deaf and insulting. More cases of Signal saying "we know better than you. You're using it wrong. Do what we say.
It was a lie before, too, as their MMS support is a raging dumpster fire and when their code pukes, it just silently eats the error message, leaving one to wonder why the message wasn't sent (or received!)
This announcement totally squares with my experience trying multiple times to fix their MMS implementation. It was at that point that I stopped using Signal for SMS, since I knew it wasn't important to them
The idea that they can't improve the UI/UX to better inform to the people who repeatedly, accidentally send insecure messages/sms (ignoring the existing words "Unsecured SMS" in the chat field, the unlocked lock near messages, the unlocked lock next to the phone, or the giant banner that occasionally drops down that tells you the % of secure messages you can be sending if you pester a contact into grabbing signal) as one of the reasons for this change is frankly bullshit.
Changing the Send button's icon to "SMS" or a color/border change ala iMessage are ideas off the top of my head and I'm sure they've got designers significantly more talented than I am that can think of better ones. We've seen very little iteration there that's indicated the significance of that problem...and frankly if they highlighted this as a tactic vs endless spam texts more people would be receptive to this news. As it stands I think this is going to significantly reduce their number of casual users. In fact I'm willing to bet that the cohort of users who are used as justification are the least likely to convince their contacts to switch to Signal.
Don't get me wrong, their real desire to increase the amount of people sending secure messages via Signal alone + resource mgmt in the face of a recession are valid. But acting as a unified messenger (with better link unfurling, threaded replies, and reactions after Google killed Allo vs the default messenger that spent years getting them) was the trojan horse onto many of my friends' and colleagues' phones. Now that there's parity I can see more people just opting into the default messenger/FB Messenger + Whatsapp combo because more people exist there and we're all just lazy.
This is sooo bad. It doesn't matter what reason they give, this is the only reason I can get some people to use Signal and it's the main reason I found it interesting in the first place.
FFS. I've been fighting uphill this past 5 years to get my colleagues, friends, and family onto Signal. This single decision will tank any hope I have of keeping them there. Literally the only reason I'm able to convince non-privacy advocate types to switch to Signal is because it is a drop-in replacement for their existing SMS application.
This decision is idiotic and will cause a mass migration off the platform. Why not take a better approach and work on a better UX to make it clearer when a message thread is secure or not?
> The most important reason for us to remove SMS support from Android is that plaintext SMS messages are inherently insecure. They leak sensitive metadata and place your data in the hands of telecommunications companies.
Ok I get that on Android the situation is such that, as a message provider, you don't give away "metadata" ie who is texting whom, keeping that data either for yourself or the highest bidder. WhatsApp, too, fuss about e2e encryption while conveniently not talking about the value of "metadata" for ad targeting and even want to aggressively grab and upload your contacts at every turn (despite it being illegal in EU to share PII without explicit and documented and revocable consent of all individual phone number holders stored in your phone book). But why does this change come only on Android? Would it be suicidal for signal to drop SMS/MMS when the default messaging app (iMessage) does fall back to SMS/MMS on iOS as is well known?
I for one welcome the change, because my phone does not have an SMS plan (data only) and the "send by SMS" is a bit confusing.
A messaging app should have one clear behavior per interface. This was "maybe secure, maybe not". I have an SMS app for that (well, VoIP-sms, because I'm weird).
You sound misguided if you're trying to explain the details of signal to get them to use it. All they need to understand to use the app is "its sms", any e2ee they benefit from as a result are completely in the background.
I remember back when I had Android, how amazed I was that I could just make Signal the new default messaging app, and don’t worry about who were Signal-users and who weren’t.
It made it amazingly easy to get started yourself, and also convert others.
Why on earth would they decide to give up that advantage?
I get their reasoning that SMS is insecure and you don't want to accidentially send an SMS. I use Signal mostly for "confidential" things, but every now and then for the occasional person who contacts me there. So Signal is my "secure" app, Whatsapp my "family" app, and so on. It's really weird if a family member shows up in my secure activism chat app.
It would make more sense if there was one codebase that supported all apps. And then I could make a "silo" for each use case. I would make one icon for activism, one for work, one for friends. The first one must use E2EE, the second one must use my company's Rocketchat, etc..
It's a pity Signal doesn't allow third party clients. I really hope somebody makes a rouge multi protocoll app, like Pidgin used to be. I bet a dedicated small team could make it in a year.
This is beyond stupid. This is the only way I was able to convince friends to use Signal. Heck, it's one of the only reasons I used it myself. Didn't have to juggle two apps.
It was obvious this was going to happen when they refused to implement RCS.
So instead of working on RCS, we got mobilecoin, stickers, gif search, and now yank out legacy SMS support so more "features" can be developed?
As an early adopter of TextSecure, through CyanogenMod integration, to Signal and everything in between, I have the t-shirts and all -- I am done with Signal.
This is really going to mess with some highly stressed out, low digital literacy people in my life. I guess I'll need to help them move to something else - is there any other basic SMS app on Android that a) looks like it's from a legitimate developer, and b) doesn't skim your message content for ad personalization?
Combining signal and sms to have a single messaging app is a big reason as of why I keep using it.
But like many recent developments, I'm just left dumbfounded by their high-level decision making. I've stopped recommending signal to tech persons for a while. I don't want yet another messaging app either. Matrix is serving me well.
This is terrible. Most of my social network is not yet on Signal, but using a single app for all my communication makes my life so much easier. Signal was always promised as the one-app that everyone could use even if their network was not using Signal.
Is anyone NOT inside Signal happy about this decision? Please comment if so, and why.
Signal has the worst product sensibility of any company I've ever seen.
Dropping the Chrome Extension was a major quality of life dip for me, made Signal far less usable across systems. Their insistence that they didnt feel it was up to their desired quality offered me no comfort; it worked, it was easy, and you tool it away.
Signal refusing to allow scripting, or an API, or any option at all for expanding user agency sucks.
They have voice messages but to my knowledge you cant preview the message at all. I dont even know if you can abort sending it once you start?
This company had such an early lead but they keep doubling dowm on the most detached, conceited ridiculous plans, few of which benefit the user. It's embarassing. It'd be so nice if there were some mandatory protocol in telephony that let people declare other systems they support; add someone's phone number and see their XMPP, email, irc, and mumble contact info. It's absurdly difficult for people to make known their contact info; Signals sms integration was a killer feature that seamlessly put them atop the telecomm heirarchy, but here the stupid fools are, killing that killer feature & what got them this marketshare.
Terrible, made my stomach sink. I got non-technical people to use Signal. They were happy for years but now they are going to be very upset by this and the problems will flow down to me.
> The most important reason for us to remove SMS support from Android is that plaintext SMS messages are inherently insecure. They leak sensitive metadata and place your data in the hands of telecommunications companies. With privacy and security at the heart of what we do, letting a deeply insecure messaging protocol have a place in the Signal interface is inconsistent with our values and with what people expect when they open Signal.
They do have a point though. SMS is insecure, unencrypted and leaks highly sensitive metadata anyway and it needed to go from Signal. You already have the system SMS app for this to use.
Outside I think mostly the US, SMS is basically only used for spam and 2fa messages. I can't remember the last time I communicated with someone via SMS to be honest.
To hear that people use it in group chats is mind boggling to me.
US is iPhone-land. iPhone users default to iMessage and "don't want another app", so SMS is still going strong, as it's only bridge between ecosystems.
Rest of the world is more diverse, so iPhone users don't get to force their default on everyone (as it's crappy if you don't have an iPhone). Also Google constantly fails to build vaiable, cross platform alternative. Therefore everyone is used to having a few apps.
Basically situation in US is what you get if you allow entire nation to be put in walled garden.
Also it's absurd, that instant messaging, that had zero meaningful innovation over last 20 years, still isn't over open protocol and we tolerate that's used by corporations to pressure customers into their ecosystems.
hahaha, that was literally the only reason why at least consider signal over other IM, now they lost it they have literally zero benefit over Element or Telegram since you will need dedicated SMS app in phone anyway
personally I jumped the boat when they made app unusable with PIN code nag screen, which they backpedaled from after uproar but it was already too late for me and my extended family where I pushed Signal, there were message delivery issues, horrible downtime in Europe because US admin was taking sleep, but the unavoidable nag screen was the last drop, the later news about shady crypto and other stuff just convinced me this app ain't worth a dime, which this SMS announcement just confirmed
if you wanna alternative IM app use Element (Matrix), unlike Signal it doesn't require phone number, it use decentralized network and you can choose from whatever app you like, never understood why IT skilled people pushed Signal after Element became already quite user friendly
This is going to seriously harm their user base. I've used signal for years, but will have to drop it with change. People aren't interested in maintaining several different messaging apps.
I think this will only affect US users, because nobody uses SMS outside the US.
And switching between apps is the expected thing to do when trying to push people to another platform,
I have telegram, signal, whatsapp and Element on my phone, this is why the new digital markets act is going to be revolutionary, especially with bridge friendly platforms like matrix.org.
I hear this a lot but this is way too caricatural.
For one, commercial services will go through SMS to contact you. Delivery people asking if my mailbox can fit parcels won’t be through Whatsapp or Messenger.
Then you’ll also want to compartmentalize and limit how some people can reach you. That means if you’re already giving them your phone number, you don’t want them on the other messaging services as well.
Life is complicated, and there will be endless use cases for the baseline, default messaging platform.
> I think this will only affect US users, because nobody uses SMS outside the US
This is not true at all, at least for Czechia. The number is going down but it's still in billions (for a country with population of ~10.5M). Quoting from the official annual report:
> In the number of SMS messages sent from mobile networks in 2018, CTU estimates - in the context of the increasing popularity of OTT messengers (e.g., WhatsApp, Facebook Messenger, Viber, etc.) - a slight decrease relative to 2017, approximately by 2% to 8.21 billion SMS messages.
I feel like I somehow caused this mess by becoming a monthly donor.
It feels like I just got my friends to put letters in envelopes instead of only using postcards. Now we all have to drive to two different post offices - one for letters and one for cards - because the original office will stop delivering cards. Everyone is just going to go back to using postcards.
>Dropping support for SMS messaging also frees up our capacity to build new features (yes, like usernames) that will ensure Signal is fresh and relevant into the future
Oh no, the "did you remember to lock the door to the auditorium" messages I send to my boss may be unencrypted. The humanity.
What do you purists use to talk to all the random people around you? Do you reply "I'm sorry, it's too insecure for me to answer that, you'll have to install Signal first" messages when more distant colleagues ask you if you're at your desk right now? Do you teach your grandmother to use WhatsApp so her birthday greetings won't be intercepted by the NSA?
(Admittedly, the email situation at work being utter trash may be coloring my opinion here.)
Whatever they prefer. Whatsapp, imessage, or regular sms (when it makes sense). Or slack honestly. I have friends and family all over the world I only push for Signal where necessary and I want it to be good for that.
This one app nonsense is asinine. You’ll never get global adoption of one thing.
I'm surprised at how much backlash there is to this.
I've had Signal since shortly after it renamed itself from TextSecure to Signal, and I never bothered using it as the default SMS/Messaging app, because back then it was a bad SMS app. It felt like it paled in comparison to what the default Android Messages app could do. I didn't want to get the false impression, either, that my chats were encrypted when they really weren't, just because they shows up in Signal.
So I kept the two separate. I assumed pretty much everyone else did the same. And yeah, there's the occasional oddity when someone texts me over SMS instead of using Signal when I know for a fact that they have both, but most of the people doing it are using iPhones, so I have to assume it's the same experience for them as well.
What's weird is that in the numerous Matrix vs Signal comments that populate Signal and Matrix submissions you rarely find SMS support as an Signal advantage over Element/Matrix.
Most people I know didn't like Signal taking over the SMS when they accidentally opted in.
Well that's probably going to suck for everyone who convinced their non-technical parents to switch to Signal.
"My bad, this easy SMS client I got you to switch to is going to stop supporting SMS, and we're going to have to export all your old texts or they'll be gone forever."
Apple does not seem to think this is a problem. Their default Message app supporting both SMS/text and iMessage. They have an opt-in to send via SMS if iMessage fails and this gives it more reliability too.
It would actually be pretty funny if Apple enabled/defaulted end-to-end encryption in iMessage used that to bash Google's green vs blue messages whining.
this is effectively going to remove signal as my messaging app of choice. i understand that messages that are not signal messages are not secure, but it is not going to be possible to convince anybody i know to download a special app if they want to talk to me. they will just send SMS, and I will have to respond via SMS, and it wont involve the signal app.
i hope they reconsider this decision; i have been using the product since textsecure and I would hate to stop doing so because they no longer support out-of-network communication.
This is extremely frustrating and lowers the chances of me ever adopting a similar non-default texting app again. This will hurt Signal and as well as poison the well for future developers.
SMS and security are simply incompatible. And either you fall into one of two groups 1. You know sms is insecure and this is a insecure method of communication 2. You think sms sent via signal is secure because it’s a “secure messenger”. It’s clear that HN users will fall into group 1, but the vast majority of people would fall into group 2. So for me this is an overall security win.
Yeah, I think this is more likely to be the case. People who don't understand encryption but used Signal as their SMS messenger were at least getting opportunistic encryption with any of their contacts who were using Signal. Now they'll probably just uninstall it (like every iOS Signal user I've ever known).
A few years ago I would have agreed, but right now Signal is doing just fine taking users from WhatsApp (FB TOS changes + ads + social group analysis)and Telegram (sketchy non-e2ee, Russian owned, based in the middle east).
What people are concerned enough about a Terms of Service change to leave Whatsapp, but struggle with the unlocked icon next to "Insecure SMS" in Signal?
What people know that Telegram isn't end-to-end encrypted, but think SMS is?
It’s a big pie, they’re also fairly different. And to be honest it’s only a matter of time before Telegram has a (public) security incident that drives much more people to E2EE messaging.
I'm curious how long it will be before that public incident - they rolled their own cryptography right? With that, I would imagine that if it hasn't been pwned yet, then there would be a disproportionate amount of people trying to break it.
Google did a security research around ssl and a crazy percentage of people think the lock icon is actually a handbag icon. The rest of the research highlighted how most users aren’t able to make informed choices, most people lack the technical basis to make those choices.
So what's the overlap of "people who care enough about e2ee for it to matter whether a message they're sending is encrypted or not" and "people who think the lock icon next to the send button in the encrypted messenging app they downloaded is a handbag"
I'm willing to wager it's not as big as you're trying to imply.
Beyond that, the minority in group 2 that use Signal are most likely to be using default settings. SMS handling is a non-default option. So you're left with a very tiny minority.
Group 1 makes up the vast vast majority of the userbase (and most likely 100% of the evangelising userbase)
(Also: if things are unclear for non-technical users, that's a UX challenge, not an absolute)
I know it is harsh to say, but whoever approved this should probably be sacked. This is really obviously a poor decision with respect to preserving/growing the userbase and will actually decrease privacy overall when fewer people are using Signal.
Many people have commented on why this is devastating news regarding future adoption of Signal. But there is a second part to the announcement that hasn't received a lot of attention yet:
> If you want to keep them, you’ll also need to export your SMS messages from Signal into that new app.
So that means my text messages will be removed from my Signal chat history? Put differently, considering how many of my contacts over the years switched between using Signal, not using Signal, and using Signal again, this means that parts of my conversations will suddenly be gone and conversations might suddenly be incoherent?
I have trouble expressing just how angry I am about this change.
This is an idiotic decision. There are real issues around improving the UX for making it more clear when a message was sent as SMS instead of being encrypted and dealing with the problem of undelivered messages because the recipient uninstalled the app, but to drop SMS support entirely instead of improving those pain points? Terrible, terrible decision.
I hope that the signal devs are looking at this thread and seriously considering reversing this horrendous decision. Just announcing this was a horrible idea, but you can at least salvage it by formally retracting it. Getting rid of SMS support would immediately and swiftly kill the app. I use Signal for the sole reason of having a secure messaging app that works with SMS. If you get rid of SMS support, you immediately kill the app. This is quite possibly the worst decision you could possibly make.
This is a really terrible idea, and the reasoning doesn't make much sense. The three reasons given are all the same reason which is "users aren't sure whether they are sending SMS or signal messages" and is purely a UX problem.
Reason 1 and 2: Users can be confused by whether a message is SMS or Signal and this is bad for security, backed with "there's only so much we can do on the design side"...really? At what point do you sacrifice the convenience of the masses in order to get through to people who don't understand that a message that says "SMS" on it is sent over SMS? Hell, turn SMS bubbles bright red or something, that would be better than removing the feature.
Reason 2: Users can be confused by whether a message is SMS or signal and thus end up incurring charges - I thought about this for 5 seconds and came up with a solution - ask the user whether they ever want to send SMS through the signal app when they first open it and respect that preference. Make it a setting. Boom, if you are using signal you know you're going over data. Or is the response to that that users are too dumb to understand it? I'm positive there is a good solution here.
This decision just doesn't make any sense and it's probably obvious to the people working on signal that it doesn't make sense, I wonder what the decision process looked like here.
Even non-technical users understand the difference between a blue and a green message on iPhone. There's no reason they could not have made a clear visual distinction with icons or messages or something.
One Signal feature that I always wanted, and will apparently never get, was the ability to send the same message via SMS & data, and have the duplicate cancel out on the other end. Service is spotty in my region, and I routinely have either cellular or data connectivity.
I initially downloaded Signal assuming it had that feature. Then they removed encrypted SMS entirely: https://signal.org/blog/goodbye-encrypted-sms/ -- I almost uninstalled the app then.
I have nothing new to add other than to simply parrot what has been said a few hundred times above.
This is a terrible decision.
It took significant effort to convert close friends and family to signal and was only palatable due to it becoming the default messaging app on Android.
Not only will I be likely to field a ton of tech support once this occurs, I'll also likely need to recommend a completely different app.
Looks like I might just bite the bullet and buy iPhones for my immediate family (without icloud imessage backup)
I'm tired of explaining to my relatives why they can send their picture to one person but not to another, or why it requires wifi for some contacts. Mixing two incompatible messaging standards communicating via two different channels in one app is confusing for many people. Sure, it also has advantages and I think you could make it work, but the app actively asking users to make it the default SMS app is not a great idea.
I read the blog post with delight. I have been waiting for this change for a long time. I opened HN comments to join in what I thought would be a celebration, and I was surprised to find dismay. It never occurred to me to encourage my friends to use Signal for its SMS compatibility, (which isn't even supported on iOS). The whole advantage of Signal is its security, and I hated having to recommend an app that is secure, except when it's not.
> "Letting a deeply insecure messaging protocol have a place in the Signal interface is inconsistent with our values and with what people expect when they open Signal"
> "We’ve heard repeatedly from people who’ve been hit with high messaging fees after assuming that the SMS messages they were sending were Signal messages"
> "We can only do so much on the design side to prevent such misunderstandings"
It sounds like they are trying to protect users from themselves.
Till now I kep't signal around despite the fact that I wasn't really getting that many messages on the app.
Now I am faced with a decision:
* Do I keep signal around, for that one to two messages a month I receive?
* Or do I get rid of it, forcing my contacts back on Whatsapp/regular SMS?
To be perfectly honest, I am thinking about just gettting rid of it. No need to keep yet another communication channel around when I can't get rid of the other ones anyways. :(
Thing is, now they've shown a willingness to make this change, so from here on out people will be worried about them trying it again at some point.
The only way they have a hope of putting this genie back in the bottle is to provide a loud, strong, clear mea culpa, stating that they were categorically wrong to propose dropping SMS support, plus a strong promise that they will continue supporting SMS for the life of the product. Maybe something along the lines of, "if we ever propose dropping SMS integration again, you can consider that a warrant-canary type of alert".
SMS/MMS needs to die at this point. I am glad to see Signal take a hard line here even though it will cause some headache for users, of which I am one, though I do not use it to send outgoing SMS/MMS.
These protocols are insecure, not private, and fundamentally incompatible with Signal's mission. Supporting them at all, while highly convenient, is a queer oxymoron for an app like Signal. We have to rip the bandaid off eventually.
The theory is good, but when my non-technical schoolteacher neighbor wants to use her iPhone to tell me that my package was delivered to her house incorrectly she's going to pop open her iMessage app and look up my phone number, and that's going to come to me as an SMS.
There's no chance I'm going to get her to install Signal, she doesn't need it, her circle is almost all blue-bubble iPhone users who don't value anything Signal adds over what iMessages gives them.
This doesn't kill SMS/MMS. Not even a little tiny bit. All it does is make MY life more irritating because I have multiple apps that I have to deal with now. The way to kill it is to make something better that people WANT to use, that offers the extra value to make it worth the effort.
What do you suggest as a replacement for SMS? The replacement has to be supported by every device everywhere (including feature phones), should not require any deals with a FAANG company, and should still work when your data plan runs out. Being unencrypted is completely unimportant for 99+% of the uses I still have for SMS, but those uses are still important.
Mmm...
You can use end-to-end encryption with Telegram if security is your main concern.
I don't see how it being "Russian owned" is of any concern, but if you feel like privacy is an issue then you're free to claim their $300,000 prize at stake or just take part of their ongoing bug bounty programme and be rewarded by spotting flaws.
Well this just killed signal for me. The only reason I got my close family switched over to signal at all was it's SMS support, now that only the family uses it none of the family will bother opening it.
This trend of siloing functions into seperate apps is confusing and frustrating for regular users, especially elderly users like my mother "Just push this button you can text and call me and text any of your friends"
It'll be interesting to see how their user numbers change.
(How many current users will it drive away? Or cause to use Signal less than before?)
(How many new users will Signal acquire, because adoption network effects weren't working as well as possible, when messaging with non-Signal friends was too convenient, but now Signal users are more motivated to prod their non-Signal friends towards Signal?)
And who's going to pick up the users that Signal loses?
I pretty much only use the signal protocol to chat with my husband, who I convinced to install the app because I could help with any problems that arose. I'm not going to use one app to communicate with just him and another to communicate with everyone else, nor is he.
The result of this change is that we will stop using signal all together. They've accomplished the exact thing they said they want to avoid.
When people complained that adding emojis and animated gifs was adding attack surface, Signal (moxie) said that it was required to build the user base, providing more security to all by providing more chaff via moving more people off SMS. And now they want to kill user engagement by forcing people to use a special app? What a boneheaded move. WhatsApp is going to pick up lots of users.
Thanks, I'm sure Signal devs to some extent look at HN, but was looking for a way to concretely let them know about the potential damage I see this causing.
I'm sad. I actually donate to signal every month, but now that will likely stop and I'll have look for alternatives.
Rip. This is definitely going to make it harder to get signal adoption. My partner will surely stop using it too now and I'll have to convince my friends to migrate to yet another platform.
Why doesn't anybody fork the Signal clients? There are so many bad design decision in the clients (for instance no message backup on iOS or no way to save all media to storage automatically) that I don't understand why people accept the Signal Foundation's stewardship of the client code.
There are forked clients, but usually you can't use Signal's server infrastructure, so you need to roll your own, and now it brings another set of problems.
Because while both the server and the client are Open Source, the server doesn't federate. If you want to be able to communicate with anyone, you have to use the official server instance. And the official server instance doesn't allow unofficial clients (though some clients seem to get away with it for a while).
Molly[1], a fork of Signal, seems to work fine. I've used it for a long time and never had any issues with it (and it connects to Signal fine). But for security reasons one of their changes was dropping SMS, so switching to it won't do you any good there.
Just to add to the frustrating elements of this shockingly bad decision:
In my friend circle, at least, it's common for people to go in and out of using Signal. They might have had it on an old phone and forget to install it on the new phone. Whatever - life happens.
Signal can't know if someone who used to have their number registered with Signal has stopped using it. Signal will still display them as a user and accept messages. It's been invaluable for me to be able, if I message a friend after a break in communication, to send a signal message...and then, if I don't get a response, a SMS message. If they respond to the SMS I can see in our history that they had signal and switched at some point. This change takes that away and will make it must more difficult to deal with inconsistent adopters.
This is a frustrating change being proposed and I don't like how powerless I feel to stop it. I even started donating to Signal because I support what they have done but it will dramatically limit the usability of the app. Many people sign up for Signal and then never check it and so it was convenient to be able to send a insecure SMS message to them instead.
The only possible benefit to this would be to break their dependence on using phone numbers as the way to sign up for accounts and possibly provide a reasonable way to export message data.
Otherwise it just feels like the wrong decision and a reminder that Signal is not a community driven project but subject to arbitrary changes and provides no way to fork or disagree with the project lead as can be done with most free and open source software.
This is the type of occurrence which leads me to refer to most user-service relationships as non-consensual.
The user enters the relationship consensually, but the choices about the service's operations are done without the user's consent.
In this case, the user's only choices are to either abandon the service, or to put up with the changes they did not consent to.
In the future, with data portability being common and table stakes for most services, I think there will be a third option: seamless transition to a different service, preserving all data, metadata, relationships, and user accounts.
This is already possible with existing, established technology: private keys, hashing, and text files.
We have a bright future to look forward to, where this type of change will be perceived as old-fashioned and barbaric as surgery without anesthetic.
> We have now reached the point where SMS support no longer makes sense
> ...
> Now, data plans are cheaper and far more ubiquitous than they were nearly a decade ago
I'm curious, are these guys lives in a bubble or what? I think they should try to travel around the world a bit.
> we’ve heard repeatedly from people who’ve been hit with high messaging fees after assuming that the SMS messages they were sending were Signal messages, only to find out that they were using SMS, and being charged by their telecom provider.
So in essence, they fuc*d up UI/UX and now the simplest approach to fix would be just to remove it. Sounds like a brilliant idea from an MBA guy or whatever-evangelist-title-is.
My god, so much good will burnt in my family and friends circle switching to Signal and now this. As a plus I went to the google play store (in Canada) to install google messages to anticipate this change on my wife's phone and the top result is 'Messages' which is a 3rd party app with ads and 1 million + installs. It also has a blue icon similar to the official Messages app. Nowhere does it say it isn't by Google until you scroll the to bottom and click on Developer contact to learn it's made by some rando @ gmail.com...
Absolutely not. Nobody wants to have to use a second app, especially after having it this way forever. Where's the change.org petetion?
I will be recommending against using Signal for any reason whatsoever to unless this decision changes. If it goes through, I'll move myself and everyone to something else. The options for e2e encryption are many today and I already have to have a bunch of these apps, so Signal becomes pointless. If they do this, they'll do worse later. Better to get out now at the "first" red flag.
First, I detest this. As an iOS user it’s annoying to have another messaging app and I’m sure many android users will stop using signal. One day I converted my whole extended family to signal by just installing signal on their android phones. Done, no change for them in their user flow.
That said, I also want to use signal without my phone. Things like usernames would be great.
That said, part of me thinks that’s an engineering problem, not a UX problem. Why are engineering problems being pushed into the UX requirements?
Terrible decision. They have a nice blue branding. What they could've done instead is to show SMS messages as green bubbles and then we'd have: green bubbles / blue bubbles, just like with iMessage, except this time it works both on iOS _and_ Android. This might win them over more and more users.
If they manage to make the UI and feature set as complete as iMessage, it would convince people to switch to Signal much much faster than Google's pity RCS bashing of Apple.
Bad move. They should be expanding support to include RCS (which can support e2ee, although I don't know if it's at the provider level or at client level).
Further — can anyone recommend a good Open Source SMS Android app? The only ones I can find are AOSP Messaging and Simple SMS Messenger, both of which are "Okay".
Signal consistently has been a poor UX for me. Sure it's super secure and that's nice. But I don't really care about the security of the convo with my aging parents. I care that they can easily respond to me.
I'm happy we have an available secure chat for people that need/want it, but I'm more than happy to keep it relegated to niche uses until it gets more user friendly.
I see a lot of pushback against this but even WhatsApp doesn't have this feature. Signal is just a small team of hackers (like 2 dozen employees) fighting against big tech (thousands of employees/developers). They aren't going to be able to support everything big tech does and what big tech doesn't. It is a pick your battle thing.
I do think Signal deserves a lot of criticism but I'm always amazed how a forum of programmers and highly tech literate users just trashes a small team of hackers fighting against big tech. They are open source. We are the ones that can help them. There are plenty of custom builds out there (that do access official Signal servers) and you can build this feature back in if you want. I don't think it is a problem if Signal decides it has more important features to support with their tiny team. But if you want more features you got to donate either time or money. This is "HACKER" news, so get hacking.
This seems opposite to what their new president said last month:
"From the beginning, the team behind Signal put people and their needs at the core of their commitments. They understood that iron-clad security is fairly pointless if people can’t use, access, or feel comfortable with it. In other words, if my friends won’t use a messaging app, it doesn’t work as a messaging app. It works as a thought experiment, at best. Understanding this, Signal’s developers and designers created an app that honors people’s needs and expectations, while maintaining strict privacy promises."[1]
I'll echo the other comments here talking about network effects, onboarding friction, and social capital wasted convincing friends/family.
I like the Signal app for my SMS messages. Almost nobody I regularly talk to uses Signal so I mostly use it for this purpose. I might as well remove it and get rid of yet another app listening for cloud notifications and draining my battery.
Maybe I'll grab the source code, rip out all the Signal parts, and just use that.
I support this decision, I don't use SMS and I'm in support of everything that kills SMS.
Next step: Please stop using phone numbers as a user ID. I have lots of throwaway phone numbers, but many people don't want to leak their phone number to every single person they want to have an encrypted conversation with.
Yeah, I'll just tell the gas company, electric company, internet provider, my bank, my elderly neighbor who can barely use a phone and I taught how to text, every restaurant I order online from, the plumber I just texted literally an hour ago due to a pipe leaking to just... not use SMS. I'm sure they'll listen.
I assume you live in a place where SMS isn't necessary? In the U.S. it is.
> my elderly neighbor who can barely use a phone and I taught how to text,
I tell them to either e-mail me or stick a handwritten note on my door. E-mail is WAY easier to use for elderly people in my experience. You get nice big keyboards, big fonts, big screens, and it works on any device you own, not just one. But if they disagree they can still handwrite a note to me
> every restaurant I order online from,
I use a fake number for these. They don't need my number any more than I need their wait staff's phone numbers. Never been a problem. I just go pick up and say my name, no SMS bullshit.
> gas company, electric company,
Especially when there are issues that's how mine send updates. To say nothing about companies that require 2FA through text!
> bank
I can't use VOIP numbers with them, not sure about Twilio.
> my elderly neighbor who can barely use a phone and I taught how to text
You make the assumption that they even have a computer: they do not. They do normally just knock on my door, but they want to send and receive pictures to their family and other people who do not live close by.
> every restaurant I order online from
I want to know when my order is ready.
> the plumber I just texted literally an hour ago
He asked for a picture of the leak and to text it to him. He's reliable and has done good work before, I'm not going to switch just because he doesn't use email.
My point in all of this is that in the U.S. SMS is ubiquitous. As much as I would love to leave it behind, there are just so many situations where you need SMS.
Honestly not really, in the US. You can usually find ways around it if you tell the business that you don't have SMS. With governments I don't think they can legally require you to have SMS.
When they find out it's incredibly difficult to deal with you because of the design choices they made, it helps dethrone SMS, one business at a time. Vote with your behavior. Make them realize they made a bad choice by picking SMS.
I hate SMS too, but I think this decision will hurt Signal infinitely more than it will hurt SMS. By that I mean it will not affect SMS at all and only Signal.
I tell everyone I don't use SMS. The only ways to message me are e-mail, Signal, WeChat, FB, and Instagram.
E-mail is the best "generic" way to reach me that isn't tied to a company's platform, and a much, much better UX than SMS in almost every way, especially when travelling internationally with multiple devices.
I haven't for many years, for sending (except for one time I wanted to test a modem driver SMS function).
I regularly use Signal, Telegram and Google chat and used to use whatsapp until it was banned by my employer but the only time I ever use SMS is to receive automatic authorisation SMSs
It's so very obvious that removing SMS support for "security reasons" is just as saying "please store your insecure SMS plaintext with your stock messages application or better just use WhatsApp for that again".
It makes me wonder if there might also be another reason. Legal issues (encryption, governments, etc)?
If not: please Signal guys - that's just a step back. People who write SMS won't stop writing SMS, you just allow another party to do whatever is good for them. And searching through _all_ messages is really a big thing for some of us.
Why not just add some awareness items - visible eyes and ears and dollar signs, together with the information what information is given away and that this might cost money?
Glad I was immediately suspicious of the sms feature and decided to not use it. Seems to be an unpopular opinion, but I'm a big fan of compartmentalization when it comes to closed ecosystems. This change won't affect me or my sms chat history.
I use Signal because it's a better SMS client and being able send/receive messages without a SIM card to my family is nice when traveling. Encryption is like a distant nice to have for me and I'm sure this is true for many others
Privacy being the primary goal of the app, they should remove the phone as username tenet. This is almost as bad as it can get for privacy, e2ee or not.
"We have now reached the point where SMS support no longer makes sense."
Totally disappointed because of this. I convinced tons of my friends and family to use signal for theur all messages and sms and now what !
Me too I use it heavily for sms.
They could just put another tab for sms and people would be happy.
This is great. People should never use the number associated to their SIM card anyway as it allows telcos and others to track location and more. You can switch your real number to a VOIP provider such as voip.ms and use Linphone or other app for SMS. Use burner sims you pay for with cash such as Mint Mobile and only activate/use them away from home network when needed.
If Signal does enable creating accounts with non-identifiable user names instead of phone number, then it will be a great improvement and a protocol that can be used for activists.
Why do I have the feeling that every person who complains about this as something of a deal breaker are from the US? This is so weird, the rest of the world moved on SMS 10-15 years ago.
I set Signal as my default messaging app until I was texted while my phone was off and the messages never showed up later. Could certainly have been a problem with my mobile service provider (Xfinity Mobile), but it's not an issue I've ever had before and seemed like an especially unsurprising result of using something other than the default messaging app. Curious if anyone else has had a similar experience
If anyone want a privacy focused all in one app I cannot recommend Beeper enough. I have been using it as my main app for SMS, messenger, WhatsApp and LinkedIn for half a year now and have only positiv experiences. Some bugs still, but amazing support and continuous fixes.
I can see how this is a hassle to maintain though; just for example, my Huawei consistently resets the default sms app to the crappy stock one every time I use their "ultra battery saver mode" (which I otherwise like a lot) even though I explicitly included signal in the list of apps that are allowed to run in that mode.
So I can see how the ecosystem makes this an annoying feature...
Imagine a world where Signal supported 3rd party apps on an open protocol.... This would be no big deal.
I find myself very glad to have focused my efforts on moving family and friends into the Matrix ecosystem instead of Signal. Not that Matrix is perfect, but I don't have to worry about the rug getting pulled out from under my feet. :(
I am not so much upset about the decision to remove SMS support, but about the reasons they give. It smells like a really lame excuse.
But whatever. I only send and receive SMS very rarely these days, so I installed Silence on my phone. It's still annoying, though. Having one app for SMS and encrypted messaging was very convenient.
For a starter, i don't rembember having a choice to replace the default messaging app with signal when i installed it. Also don't remember any settings to split it back to default messaging. It has always been All-In-Signal or no-Signal-at-all, and no user choice. Am-I wrong here?
How is it a serious UX/design problem? iMessage just makes SMS messages green and it's so effective at conveying the difference that people claim it creates social stigma against Android users.
In the long list of SMS alternatives below, can someone tell me what's wrong with the default Android SMS? I use Signal and regular SMS, why would I install a second SMS option for non-Signal ?
I'm happy with the change, SMS to should be sent through native SMS app, while Signal is just another chat client. Never understood why they have decided to overtake the default SMS app.
Amazing decisión signal. I always hated this combination as it confused me constantly. I’ve been coding for 30 years and I’m a published author in security… and it still confused me.
I just wish there were some way to back up my messages on iOS. Their built in transfer doesn’t work. Luckily it didn’t delete all my messages like it did to my friend :/
IMO this is good. Blocking users from RCS, [while obviously no where near as good as signal, is still far better than SMS] has been a flaw in how signal has handled things
This sucks. I get the decision from a development pov but from a user pov it's awful. Having 2 apps for texting is not great and ultimately only creates friction.
I've only found only one good option to unify messaging on android. Blackberry Hub will bring together SMS, WhatsApp, Signal, multiple emails, Instagram, etc.
In the US? Old people (so, probably some members of your family), spammers, and ~100% of businesses that communicate via any kind of IM service instead of or in addition to phone and email.
Oof. As an Android user, this sucks. Though I have my frustrations with Signal (cellphone number, address book hashing, centralization, the cryptocurrency stuff, removing storage encryption) -- it's still the only app I trust. Even more than the stock Samsung messaging app. I don't want to trust another, and I don't want to have to bifurcate my messaging flow.
All of my family use iOS though, so this is already their use case. I understand less code is more secure, and a unified codebase between devices is good -- heck. This might even lead to no more phone number requirement.
But this still stinks for my use case.
FWIW though, I was more upset about the cryptocurrency thing.
As an android user myself, I much prefer having SMS built in because I use the search feature often to look back through all my SMS/Signal chats. I also regularly forward an SMS message to a Signal user, or vice versa. I'm already starting to feel like those iOS users who told me "I don't want another app"...
Signal seems to be trying to move further and further from "my preferred way to chat with people" and closer to the chat equivalent of "that protonmail account I only log in to when I need secrecy".
I obviously love having security on messages in transit, but I also like being able to keep my message history around and search my conversations for something that happened a year ago. It seems like Signal is on a trajectory to turn everything into disappearing messages. Are they the "safe for activists" communication app, or the "let's try to make as many as possible safer by default" app? Feels like they don't know.
And on top of it all the messaging is just frustrating. "we've taken away an incredibly useful and heavily used feature so we have development resource to better implement shitcoins and such" is such an irritating defense of the decision that I disabled my monthly donation.