Hacker Newsnew | past | comments | ask | show | jobs | submit | more PeterBarrett's commentslogin

One of our systems runs AWS code repository in parallel to Github and builds are triggered from there (but not in us-east-1). Time to migrate the rest of our systems to having that fallback.


SQS went down for us in us-east-1 and we lost health checks on instances there. Fully recovered now.


Well they're not selling access to data, the data is freely available. They're selling a service that packages that data for you.


Data which probably was obtained through the API which from my interpretation of https://developer.twitter.com/en/developer-terms/agreement-a... is not permitted.


If obtained through scraping vs the API it also isn't allowed via the regular ToS https://twitter.com/en/tos and their robots.txt

So either way sounds like a ToS violation. Which Twitter is usually pretty aggressive about shutting down.


I can’t see how this is true as there are tons of companies selling access to data obtained from the Twitter API such as social media monitoring companies (Sprinklr, Meltwater)


Those companies have business development deals in place with Twitter that authorize them to do this.


There are plenty of other non-enterprisey companies that sell Twitter data as well*

*I was a founder of one of them.


Interesting. Do you have any advice on how I can try this again without getting suspended?


If you are creating a Twitter app, and letting users login via Twitter and then using those user tokens to make API calls, that’s a good first step, I am not sure if you are doing this already.

Other than that, I would just make sure you position your service as something that benefits Twitter, describe it as something that allows you to create better tweets or engage with your followers better. The internet is filled with apps that sell Twitter data like Mention.com, SocialRank, and Social Animal.


Thank you this is really useful info. I am indeed doing as you say, and I think you're right it's the positioning that has got the app in trouble.


As a counterpoint, stock data is freely available (through online websites, your broker/app, etc), yet it costs thousands of dollars per month to actually license and acquire such data through the proper channels. Trying to build a business based on scraping such data (or even using your broker's free personal-use API) will get you in trouble pretty quick by the exchanges.


That's true in a sense. But I would still be paying them money to receive data from Twitter. So in a sense, yes, I am paying for Twitter data. In any case, I have a hard time seeing how this wouldn't violate the ToS.


You can’t package and redistribute tweets only tweet ids.


I hope you stay being the only person who has said that, 1 region being gone is enough for me!


I should have said, only the Console and CLI was down for us, our services remained up!


Console has lots of us-east-1 dependencies.


Someone did a really funny presentation on all apps moving to the web and then eventually OS's, followed by OS's running on OS's on the web. If anyone know's where I can find it I would really appreciate a link!

It's pretty relevant to this seeing as it does seem to be where we are going.


I believe you are looking for Gary Bernhardt's talk:

https://www.destroyallsoftware.com/talks/the-birth-and-death...


Fantastic! Thanks for that, I was searching for a good 20 mins, much appreciated.


A fun program for the sake of fun, so nice to see!


Looks like it's just guessing the object and then finding images tagged with that object. I was hoping for some sort of scene detection to find specific scenes from films.


Ah shucks! Well, you saved me a click.

Was hoping for some AI visual processing thing and was doing the mental gymnastics of how to store all that data.


I can't see how that's relevant to him asking for feedback, you're being needlessly pedantic.


My motivation to ask is:

1: I really want to understand it. Why does someone say they "finally finished their first MVP" when they already have been doing Show HNs for similar projects for years?

2: I am a bit afraid that we will see an arms race of emotionalized Show HNs if it turns out that these get more upvotes. I wonder if this site would have gotten as many upvotes if it left out the "9-5 job and 2 kids" part. I think it is worth asking if we should allow sensationalized / emotionalized titles for Show HNs at all.


FWIW, I had the same reaction. The post title does literally nothing to describe the actual project, only some tangentially related circumstances under which the thing was developed.


Welcome to Hacker News


This is so nicely written, good job.

It's a topic that captivates me every couple of months, most recently I've been interested in the power of YouTube's recommendation system. If you show interest in a topic you're slowly guided into more videos on the topic, to the point where you're more likely to buy items surrounding the topic you have become interested in. Then eventually the recommendations change, your interest in that topic weans and you're on to your next interest. It's a cycle that's easy to fall into and easy to break if you notice, but how many people don't notice?


YouTube recommendations are awful. I watch one video out of curiosity and am then flooded with similar videos. There is no nuance just a deluge of shit. Even stuff I am interested in is annoying to see day in day out. It pares the world down into a few niches in turn stopping you from discovering new and interesting things outside of what you already know.


Do you mostly watch videos from people you already subscribe to?

YouTube behaves the same way for me, and I am wondering if it's just a perception thing. If I watch an MTB video, it's mostly MTB recommendations, but they're all people I am subscribed to so it's not notable. But if I watch a video about metalworking, it probably shows the same kind of spread just for a topic and channels I'm not familiar with, making it feel like it's lost it's way.

I also wonder if they don't weight subscriptions that highly for the recommendations, and because I mostly watch subscriptions, if I watch a new topic it's heavily weighted because there's not much other data.


Yes I do and I also think this may be part of the issue.


That is a very interesting perspective. Maybe that is why I don't mind watching clips from Apples yearly shows, which are 100% commercials, but loath youtube ads/sponsorships. They are both advertising, but one is really high quality, the other is squarespace.


YouTube refs are great for music, but awful for any kind of ‘content’


It probably works on most people but I find that YouTube recommendations are always something I do not want to watch. Now a days they have started putting mainstream media and news content always at the top in the recommendations on the right side of a video and I never ever click on them. I have tried the “Don’t show me this” but it doesn’t seem to work. It’s gotten to the point where I might look into using content blocker to block that HTML element all together.


We have found that refusing to support IE11 in healthcare environments has lost us 0 customers and given them a reason to move away from it. At this stage any company using it and the likes windows xp or windows 7 are just creating unnecessary security risks.


I think at this point it would be some US auto-body shop with IE6 hooked up to some internal software, or maybe one of those big Japanese companies that never changes anything at all.


Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: