We use GitLab on the daily. Roughly 200 repos pushing to ~20 on any given day. There have been a few small, unpublished outages that we determined were server side since we have a geo-distributed team, but as a platform seems far more stable than 5-6 years ago.
My only real current complaint is that the webhooks that are supposed to fire in repo activity have been a little flaky for us over the past 6-8 months. We have a pretty robust chatops system in play, so these things are highly noticeable to our team. It’s generally consistent, but we’ve had hooks fail to post to our systems on a few different occasions which forced us to chase up threads until we determined our operator ingestion service never even received the hooks.
FWIW, GitHub is also unreliable with webhooks. Many recent GH outages have affected webhooks.
They are pretty good, in my experience, at *eventually* delivering all updates. The outages take the form of a "pause" in delivery, every so often... maybe once every 5 weeks?
Usually the outages are pretty brief but sometimes it can be up to a few hours. Basically I'm unaware of any provider whose webhooks are as reliable as their primary API. If you're obsessive about maintaining SLAs around timely state, you can't really get around maintaining some sort of fall-back poll.
> you can't really get around maintaining some sort of fall-back poll.
This has been my experience with GitHub Actions as well, which I imagine rely on the same underlying event system as webhooks.
Every so often, an Action will not be triggered or otherwise go into the void. So for Actions that trigger on push, I usually just add a cron schedule to them as well.
Completely agree on all points. We've had dual remotes running on a few high traffic repos pushing to both GitLab and GitHub simultaneously as a debug mechanism and our experiences mirror yours.
Jesus, that's a staggering figure to me coming from senior developers. I guess I'm the odd one out here, but ChatGPT is nothing more than an index of Stack Overflow (and friends) for me. It's essentially replaced Googling, but once I get the answer I need I'm still just slinging code like an asshole. Copying the output wholesale from any of these LLMs just seems crazy to me.
If you’re using ChatGPT directly for work then I believe that you are doing it so profoundly wrong, at this point, that you’re going to make really incorrect assumptions.
As we have all observed, the models get things wrong, and if you’re wrong 5% of the time, then ten edits in you’re at 60-40. So you need to run them in a loop where they’re constantly sanity checking themselves—-linting, styling, typing and testing. In other words; calling tools in a loop. Agents are so much better than any other approach it’s comical precisely because they’re scaffolding to let models self-correct.
This is likely somewhat domain-specific; I can’t imagine the models are that great at domains they haven’t seen much code in, so they probably suck at HFT infrastructure for example, though they are decent at reading docs by this point. There’s also a lot of skill in setting up the right documentation, testing structure, interfaces, etc etc etc to make the agents more reliable and productive (fringe benefit; your LLM-wielding colleagues actually write docs now, even if they’re full of em-dashes and emoji). You also need to be willing to let it write a bunch of code, look at it, work out why it’s structurally deficient, throw it away, and build the structure you want to guide it - but the typing is essentially free, so that’s tractable. Don’t view it as bad code, view it as a useful null result.
But if you’re not using Claude Code or Codex or Roo or relatives, you’re living in an entirely different world to the people who have gone messianic about these things.
Yes, I was using "ChatGPT" colloquially. I’ve tried Claude Code, Copilot - I currently have Gemini running though madox2's vim-ai plug - and a few of the others, but I’ve never personally seen the kind of 10x productivity gains some people here talk about. In my experience, it’s typically been quicker to just write the code myself than to wrangle with the model’s output, iterate, and debug. That’s not to say the tools are useless - like I said I think they're great at exposing docs that are disparate or obtuse (looking at you Google) but for actual problem-solving and building, they’ve unfortunately rarely felt like a net win.
I'm sure you're correct that a lot of this comes down to workflow and domain - in mine, the overhead of prompting, reviewing, and correcting usually outweighs the benefits. Of course, as always, this is just one asshole's opinion!
That's not actually the only issue, unfortunately. Steam did in fact accept Bitcoin for a while, but it was dropped in 2017, with Newell stating:
> “We had problems when we started accepting cryptocurrencies as a payment option. 50% of those transactions were fraudulent, which is a mind-boggling number. These were customers we didn’t want to have.” [1]
That's quite interesting, I wonder how they were fraudulent as bitcoin transactions dont have chargebacks etc. In the article they also mention that the price for the game would fluctuate which seems to indicate they actually priced the games in Bitcoin. Normally price would be Euro, USD etc and price conversion to Bitcoin would be done at check out. I guess these were the early days though =)
Seems like a catch-22. Crypto can’t be used like legitimate currency because it’s not used like legitimate currency, and fraud can’t be addressed because there’s no regulation and no mechanisms by which to make victims whole.
There are parts of me that can see some appeal in cryptocurrency but I can’t see any way around this, at least not without impacting what made it appealing in the first place.
I think this is the biggest problem with crypto that it will never be able to solve. Every year there are millions of dollars lost in crypto due to scams (the "send me crypto and I'll give you money" kind). The instant you have a digital good people can actually buy with crypto and sell for real money, it becomes a money laundering machine.
Like you could get money from ransomware, buy a ton of games on steam, then sell the account to a third party.
This will probably be "solved" by legally exempting stablecoins from nemo dat. Basically stablecoins will be declared clean by law so recipients don't care.
While volatility indeed can be a problem (depending on how the whole thing is setup), I fail to come up with scenario of 50% fraudulent transactions... Does anyone have links to more details on what kind of fraud they meant?
Accepting bitcoin payments with 0 confirmations? It shows a complete lack of understanding of Bitcoin and cryptocurrency in general, especially in 2017 when PoW was the main driver and PoS chains weren't really mainstream.
It’s not a big deal in case of Steam – they can just pull the games from your library if the payment doesn’t go through. A bit trickier with in-game purchases (but in that case they could wait for the confirmations).
Fraudlent is accoriding to the law. If I steal your bitcoin I spend them on Steam it is fraudlent. I know cryptobros think stealing bitcoins are OK and shouldnt count as fraud but law isnt written by cryptobros.
It’s not this. Steam accepted 0confirmations to ease the payment flow. Before the transaction would be mined into a block, the fraudster would submit a new higher priority transaction sending the payment to themselves.
Having very recently gone through this process for both a “local” number A2P registration as well as a toll-free registration process I can say, unequivocally, that’s already what it is.
There is no step in the process that would deter any company from taking the ~30 minutes of effort to get the registration details in order. We run a legitimate company that was only seeking to send transactional notifications that our customers specifically requested, and it took 4 rounds of back and forth to get approved. However, each step provided a very clear error code with very specific remedies (e.g. Opt-in form must include language which states…), which would make it a rote process for automation or even manual repetition. In fact, the only thing that I could see which would deter a company which seeks to rely on spam for profit would be if the registration fees were larger than their expected profit (they’re not, thankfully).
I don’t even mean this comment to necessarily be an indictment of the process - it was…fine? I’m just not entirely sure what it’s purpose could be except to collect additional fees for each SMS campaign.
That's like asking why a weather app needs both a date and city - the temperature on July 4th matters, but it's completely different in New York vs. Cairo. Same with this - the year sets the global context, but the location defines the local experience.
You and I were seemingly writing our comments at the same time - and yours describes what I was trying to say so much more successfully than my own! Completely agree on all counts.
It’s funny you mention that. I moved to Melbourne AU a while back and quickly realized how bad the coffee I had been drinking my whole life had been. I honestly had trouble finding a bad cup of coffee there - although as it’s primarily an espresso based coffee culture it’s admittedly quite a different animal. I dove in pretty deep over the years, finally ending up with a Silvia/Mazzer setup, but oddly enough would sometimes find myself longing for a pot of “shitty diner coffee” - particularly on the weekends. After being introduce to real American diners, my wife - a lifelong Australian - also occasionally has the same craving. We always look forward to our first diner breakfast whenever we had back stateside. I guess at this point I just classify it as a different beverage altogether!
There’s something especial about crappy diner/gas station coffee. I can’t enjoy American diner breakfast (eggs, bacon, toast) without crappy drip coffee. It’s essential.
McDonalds has surprisingly “good” coffee in this regard.
I hear you on diners coffee and the fact that it feels like it a different beverage altogether. I also crave it from time to time.
—do you have tips on how to actually brew it in a home setting? Or this is something can only be achieved by brewing large batches of the stuff, keeping it warm somehow, and letting it go stale for a few hours?
Yes, the rise of batch brew in Melbourne has scared me a bit - because while it is generally higher quality than most American diners at the end of the day it’s just a large pot of pour over that’s been sitting in a carafe. When there is so much good coffee to be had the city it seems criminal! (Yes, I’m being hyperbolic)
Sorry, but exactly what point are you trying to make here? Are you suggesting that NPR has never interviewed - say - Christian fundamentalists (they have)? Are you suggesting that they should interview more of them? What, precisely, would make you happy here?
As I've been told for the last decade, "everything is political" therefore NPR can't provide unbiased or neutral coverage of anything, therefore there should be no federal funding of NPR or PBS. Ideologues and corpirations donate more than enough money to sustain both without the pretense of impartiality provided by federal funding.
If "everything is political", then eliminating federal funding from NPR and PBS doesn’t solve the problem - it guarantees that only corporate and ideological interests shape the narrative. Public funding exists not to claim perfect neutrality, but to create a space where journalism isn’t entirely driven by profit motives or partisan agendas. Strip that away, and you’re not removing bias - you’re institutionalizing it.
Journalism, and specifically NPR, is already driven entirely by corporate and ideological interests. Your supposition that federal funding helps remove bias is trivially disproven by the last decade of coverage of NPR, where I literally (literally!) have not been able to turn it on without race, gender, or Trump being mentioned within a minute (it became a game).
To be fair, there was one exception. and that was a replay of a David Foster Wallace interview from 2003. Which was immediately followed by a current interview with two women talking about white men's obsession with Infinite Jest and how their podcast was helping deconstruct toxic masculinity or something like that. The comparison in quality was stark.
The time for caring about and preserving civic-level notions of neutrality and objectivity was a decade ago. I don't care anymore. If wingnuts want to unduly influence Americans through broadcasting, they can do it like everyone else--without taxpayer dollars.
If your position is "I don't care anymore", then you're not making a principled argument - you're venting. That's fine, just don't pretend it's a policy stance.
"When I am weaker than you, I ask you for Freedom because that is according to your principles; when I am stronger than you, I take away your freedom because that is according to my principles." likewise isn't a policy stance, it's naked hypocrisy.
It's a free country so people are afforded the right to be hypocrites, but nobody is entitled to receive public funding when doing so.
You're not actually critiquing hypocrisy - you're just deciding whose version of it gets a microphone. Pulling public funding doesn't eliminate bias, it just ensures the only voices left are the ones with capital to shout the loudest.
My only real current complaint is that the webhooks that are supposed to fire in repo activity have been a little flaky for us over the past 6-8 months. We have a pretty robust chatops system in play, so these things are highly noticeable to our team. It’s generally consistent, but we’ve had hooks fail to post to our systems on a few different occasions which forced us to chase up threads until we determined our operator ingestion service never even received the hooks.
That aside, we’re relatively happy customers.
reply