I mean it sounds nice but they never say _why_. "AI removes a lot of the burden" is not a _why_. I was hoping to read about some of the shortcomings of React Native, but it seems like they're doing it just because they can. I remember Airbnb had a post about switching off of React Native years ago that actually had some solid reasoning behind it.
I think we need to have a little more sympathy for GitHub. You could justify the jabs when we could all blame any outage on the migration to Azure, but then they shared numbers around the scale they're dealing with now that everyone is constantly building and pushing with AI.
I think it's commendable that they're not limiting access to the site or (intentionally) throttling newcomers. Yes, they need to get this figured out, but a little sympathy goes a long way. I personally wish them the best and hope their on-call people can go back to getting normal amounts of sleep soon.
I think somewhere down the line you have an answer to problem itself. Its for a while that companies are not into the business of sympathies, we should probably not get into that side for the topic.
1. Github has enterprise users who paid for the service, their day job requires Github to be available and working
2. Github has generous free tier which is the one which is exploring a lot more with the AI generated code.
It is a complexity in itself but the traffic should have been separated, the free users should not be allowed to bring down Github for enterprise customers (Just to clarify, I am free user myself). And if they do not have capacity it would have been perfectly fine to push back or throttle new users/repositories.
I agree that they should reserve capacity for enterprises and paying users.
But I think they’re in a tough spot. GitHub has historically been a huge supporter, proponent, and provider for open source projects. Engineers are difficult customers, to say the least, and the community would likely freak tf out of the segmented traffic.
The logical, pragmatic, and justifiable answer doesn’t always align with your market.
Not sure if I would agree to that completely. Free tier is always ripe for abuse so in the beginning for a long time, Github has this requirement that you can only have a public repository if you are a non-paying user. Somewhere down the road when machine learning picked up (everyone was on the band wagon of hoarding the data, datalakes, hadoop, flink were at the top) they relaxed that condition allowing one to have a private repository on the free tier.
They are in a tough spot cause they want to continuously ingest all the data to train their LLMs. Any fork or decentralization at large scale of git is going to impact the training pipeline.
Tbh, timing wise it appeared to me that GH started allowing private repos when one of the competitors started to allow it for free. I forget the name of the back then still competitor.. it worked on me, I retuned to GH back then..
What we are seeing here is a trillion-dollar company screwing over its existing well-established customer base in a desperate attempt to get some crumbs of the current hype cycle. GitHub isn't "dealing with" scaling issues, they are intentionally causing it by shoveling giant piles of coal onto the AI fire.
GitHub's issues would be over tomorrow if they were to put reasonable rate limits on their services. Stop the bots and GitHub is no longer permanently overloaded. It is an intentional business decision not to do this, and instead welcome bots with open arms - to the detriment of their existing paying customers.
I do have some sympathy for the low-level GitHub staff who have to deal with this - especially those who aren't being compensated with insane Bay Area salaries.
One, logically, your statement doesn't follow. They're two different companies that have primarily in different market segment.
Two, it's not true in the first place - did we side with apple and its cut of the app store revenue for instance?
They decided they needed to capture the whole open source ecosystem by turning open source work into social networking... on a proprietary platform (because open source is great, especially when it's others' software). That was before they joined Microsoft.
And then Microsoft pushed AI everywhere, including on GitHub itself with copilot.
I would have liked if they had left the open source projects alone and didn't create that FOMO for not using them.
> And then Microsoft pushed AI everywhere, including on GitHub itself with copilot.
Ding ding ding, we have a winner. I like AI. I work for an AI company. Still, Microsoft aggressively pushed GitHub users toward Copilot. They don't get to do that and complain about increased volume from AI-generated changes.
No Copilot + reasonable operation: the way things were
Copilot + reasonable operation: Well done!
No Copilot + being overwhelmed by AI commits: Sympathy.
Copilot + being overwhelmed by AI commits: "Where did that petard come from that's hoisting us?"
> have a little more sympathy for the billion dollar company
No.
Their leadership went all in on AI and in the last blog post essentially admitted some missing test coverage for a critical path. Time to learn lessons and fix your vibe coded shitslop and stop using "user graph go up" as some kind of excuse.
Our company pays for GitHub. We're paying for a broken product that stops our work. I don't have sympathy for the companies whom I pay for a product and give me broken shit in return. This is entirely preventable and their own fault.
A restaurant makes pizzas. They suddenly get 100x more popular. They can't make 100x more pizzas. But they are still taking orders from 100x more people. Not only are they not getting enough pizzas delivered that they took orders for, but in their rush to make and deliver more pizzas, they set the kitchen on fire, which makes an even longer wait for pizzas.
When the pizza you ordered doesn't get delivered, do you have sympathy for the restaurant? Or do you tell them to stop taking orders they can't fill and try not to set the kitchen on fire?
Now consider the pizza restaurant has 21 billion dollars in cash, is taking your money, and not giving you pizza.
It's a shame that the GitHub org that we use at my job that we pay a lot of money for gets affected the same way my personal nonsense does.
I don't know the architecture or any of that, but I feel like there could be (and it's not like they would've really known this until the last year or two with the massive spike) separate infrastructure for paid users/orgs vs free the same way they make the distinction with enterprise.
I get the massive load changes that they are under over the last two years, but why does a bunch of vibe coded slop take down the same resources that my company pays for every single month and has for years? I imagine properly splitting that out would be an absolute headache and not worthwhile for them vs stabilizing the rest of the service, but damn it sucks when I get blocked at work because GH is down.
Our GitlabCE instance has been sitting in the racks for nearly two years with almost 100% uptime running on 10 year old xeons that have long since paid for themselves.
Of course it is not free of all management but for our use case it is working.
There are hiccups with the CI runners from time time but nothing major and we have another machine in another rack that serves as a backup which can be brought up ~< 20 minutes.
I know companies have long since tossed their expertise for hosting their own stuff in favour of SaaS but at some point its hard to beat the up time of a single machine.
You can pay for GitHub Enterprise Cloud which is essentially their private GitHub deployment stack that they'll host for you. It's theoretically isolated, supposed to be completely separate from these outages.
It's a lot more expensive and has a good bit more limitations to the regular SaaS product.
I have a lot of sympathy for the people working at GitHub trying to keep it going, but I do not have sympathy for GitHub. That's what they're paid for. I have things to get done, and it has no sympathy for me.
Yeah, all those 1100 commits per second. Like, is this the standard for 'scale of traffic' that we're impressed by these days? And it's not like this snuck up on them. The graph shows a very consistent trend that gave them months and months to respond.
Both Microsoft and GH issue is capacity which is being solved with huge capex in DC investment, you can't create a new DC overnight and its even harder in the current climate where everyone assumes the DC is for AI and not just normal workload compute like what GH needs...
If you do the math, their numbers are something like 1000 commits per second. That's not pushes per second either, it's just the total number of commits summed up over a week, not all of which trigger much of anything to happen.
Yes, that is scale. And yes, that's not actual requests per second. But it's the sort of scale that big (and even mid-sized) tech has known how to deal with for decades. Microsoft doesn't have an excuse.
idk man, we pay stupid amounts of money to microsoft, we are an enterprise customer, expecting better availablity compared to my laptop isnt really a high bar.
> You could justify the jabs when we could all blame any outage on the migration to Azure, but then they shared numbers around the scale they're dealing with now that everyone is constantly building and pushing with AI.
Does nobody use GitLab or Bitbucket or whatever? Why aren't those full of high profile outages as much?
> I think it's commendable that they're not limiting access to the site or (intentionally) throttling newcomers.
I'd kind of expect that for all non-paid members will get rate limited in some ways, sooner or later. Same as what happened with DockerHub, they're probably just trying to hold out due to the reputational hit such a change would carry.
> they're not limiting access to the site or (intentionally) throttling newcomers
They can simply throttle access for everyone to a push/merge frequency which was normal for human-maintained projects.
They should look at how MMOs battle cheating/botting and apply similar countermeasures (it's actually quite absurd that MMOs which are actively attacked every minute have a better uptimes than Github, and on Github the problems are even self-inflicted, not caused by external attacks).
But of course throttling bots to human-level push/merge frequency wouldn't be compatible with the Github leadership's 'AI bullishness'.
For that matter, I admire their relative transparency about their incidents. I can think of other big players who will cheerfully show green statuses across the board while everyone can see that their pants are down...
I appreciate it’s tough for the SREs, but GH as a business assumes they must just fall over and keep scaling.
What if they figured out what their platform can actually handle, provided the contractual level of service to their paying customers first, and put some god damn limits on usage.
Like they should just stop signups if they can’t handle the capacity yet. It’s a business decision, not a technical decision, to decide to screw over customers instead of adding limits based on your actual capacity.
GitHub literally went all in on doing almost nothing but AI-related feature additions for Copilot. So they built their own prison.
I don't have sympathy for them. They have disrupted my work so much in the past month that it's ridiculous, and my company likely pays them millions of dollars. And you can see from this site's plotting that there is a trend towards more frequent and more critical outages. It has been extremely bad the past month.
I don't know enough about Github or Microsoft to say anything intelligent, but I agree with your sentiment in general. And, one thing is for sure - I have used Github for years, pushing many thousands of commits on personal, private projects. I've burned through as many free action minutes as I can, and I use my own action runners as well. I've done this all for free, so on balance I am grateful for them.
isn't github scaling pretty embarrassingly parallel tho? doesn't "ready to scale" at any "good" tech company mean "we can handle 10x without architectural changes?"
I’ve noticed there is 5000 api call limit on gh cli. Haven’t looked at whether it was already there or when does it reset etc. good for them finally if they weren’t already doing that.
I strongly suspect that quite a few HN members are busy creating agents that are furiously stuffing GH.
As I said before (and it was not a popular comment), it's easy to be an "armchair quarterback," with these services, as I think the brittleness was already baked in, and just waiting for the right time to crack. The only true way to have a robust platform, is to design something that will scale, from the start, and many startups don't do that, because they are feverishly trying to get out an MVP; even if it is a mess of bubblegum and baling wire.
They always say "We'll get it done right, once we get funding," but that never happens.
But my sympathy is limited by the fact that MS paid a lot of money for this, quite a while ago, and that is one company that knows all about issues of scale. They should have seen this coming.
also this site and the news's measure of github's single 9 availability (1 service at all down even copilot) is a bit harsh https://mrshu.github.io/github-statuses/
Give me a break. Sympathy? For microsoft? That might have flown when github was like seven people, but they have nearly unlimited resources to make it better. They're just choosing not to. Let's talk contracts and money before we pull the sympathy card.
I used to be on-call in a high-traffic environment where single customers pushed more bits than entire nations. I chose the role. I didn't want people's sympathy, if anything, I wanted them to complain to management.
If it gets too bad they can quit. Maybe that would be for the best, just wear the thing down until it outright fails and no one wants to touch it. One less bullshit service sucking all of the oxygen out.
> You could justify the jabs when we could all blame any outage on the migration to Azure, but then they shared numbers around the scale they're dealing with now that everyone is constantly building and pushing with AI.
How about no? They are owned by a company that is neck deep in foisting AI into everything, AI that a lot of us don't want in everything, they are reaping what they seem determined to sow.
Sympathy? It's a Microsoft company that is being ran with a consistency of a startup in early seed rounds. Their downtime is abhorrent and unacceptable as far as enterprise goes. Their engineers look like absolute amateurs allowing for such low class work it results in their customers experiencing industry leading downtime.
This is uncharitable, rude, and pretty baseless. Unless you have a lot of direct, personal information about GitHub engineers, they’re dealing with a huge spike in traffic.
Is their uptime acceptable? No. But personal attacks aren’t necessary or constructive.
Please, read between the lines. Github actions has been a mess since 2019 at least. None of the instability is new, before this unprecedented growth (for a service that's supposed to scale horizontally) the excuse du jour was the azure migration, before that it was the high rate of shipping post acquisition.
Core parts of the product, like navigating to individual files in a code review, are broken
> Core parts of the product, like navigating to individual files in a code review, are broken
I think this is a good argument to underline "It's not _just_ the scale".
Adding to this, the Github Code Review experience is kind-of broken, the way comments/threads are stacked in the PR overview has not improved, pagination isn't really a thing, and these issues are age old.
Hopefully, one day, Github will mature.
The phrasing could have been better, but we've talked as a team about cancelling a $100k/yr contract nearly half a dozen times the last year, and the only thing keeping us from doing so is various "compliance" issues, and to a lesser degree the friction from physically moving. That's a tenuous moat, and if I were a GH PM I'd probably want to know that the current instability is somewhere near critical mass.
For CI there are a number of drop-in commercially-available options. You can make it a staged migration, first trivially migrating the build runners, followed by the more complicated integration test and deployment runners. The CI harness itself can follow, and finally moving to a code hosting and visualization service is last.
The final step is challenging; likely the most difficult part is changing all the code references and imports. Shadowing changes would be straightforward. Training your likely 25-50 engineers to use the new code review UX would likely not take that long.
Considering the wasted engineering velocity during Github outages, it's worthwhile to do even a partial migration. Github's action runners have in my experience, been the most fragile part of the platform. Given the ease of moving build and merge queue runners to alternates, it's a no-brainer.
You are certainly free to attempt to migrate your workloads to your own or leased data center space in your area then buy your own on prem cloud like Dell Private Cloud and deal with all of the maintenance yourself. I don’t think it would amount to $100k/yr and you would be taking on all the risk.
For most teams selfhosting on 1U is more than enough which would cost about 1200€/year for housing + let's say 1800€/year for hardware acquisition/depreciation (probably more like 300€/year if you can work with second hand hardware) so that's about 3000€/year or 97% reduction.
Of course you may not want to manage the hardware yourself in which case you'd go for a dedicated server from a reputable provider (i.e. not microsoft/google) like OVH's RISE-XL, Hetzner's EX131, Scaleway's Core-9-L (or an equivalent solution from a smaller provider) for about 4000€/year.
Now let's say you add 8k€/year donations to your distro of choice, forge of choice, and other FLOSS projects and 8k€/year for backups. You got the whole thing running sustainably with >99.9% uptime for 20k€/year or 20% of the original bill.
Is human programming going anywhere? I mean sure you’re not literally typing the actual lines of code, but we haven’t done assembly since compilers and transpilers exist.
Personally, all AI has done is add another layer of abstraction where “pseudo code” can be the source input and out comes the compiled program. You still need to think about the design and maintenance.
No, you hire more of them, you give them what they need to do their jobs, and you actually listen to their recommendations. It's not even complicated. It just takes time and money, neither of which companies want to spend on reliability until their customers scream very loud en masse.
Typically when a company gets acquired, they get thrown a bunch of terrible initiatives like “change your cloud provider” that distract from their mission, executives saying “more AI!” and they definitely don’t hear anyone in management say “prioritize technical debt”.
If I own a restaurant and buy bread from a supplier — BreadHub.
And 99% of the bread that I get is good but 1% of the loaves, they forgot to add flour. Consistently, for years, they always have loaves missing a key ingredient that I still end up paying for.
I can be pretty sure that BreadHub have a pretty major internal issue, and should probably be questioning their competence, regardless of their “scale”, and without any knowledge of the “problems they’re solving”
I’m not saying users don’t have a right to be pissed or aren’t justified in looking at other options.
I’m saying that GH is operating at a huge scale with (probably) lots of technical debt and a forced migration to new infrastructure.
I would be (and am) highly critical of leadership. I’m not going to make strong assertions about ICs without knowing their context. I’ve worked at a company with a sterling reputation for engineering excellence where brilliant ICs were kneecapped by poor leadership.
I think a lot of us, at one point or another in our careers, have worked with potato leadership that can be short-sighted or political. It isn’t a comment on the engineers.
I mean, the leadership brought it on themselves. For a few years they encouraged sloperating Copilot. Before that, they were encouraging Github Actions usage.
Now the scale problems they're facing is because of those 2 actions of theirs. They didn't have to encourage such growth to turn a profit, and they still don't have to provide gratis services to slopware bots providing 0 value for society.
I don't think this is an apt analogy. The bread makeup is still there, as far as I'm aware, no users have lost any data or are missing "key" ingredients.
An outage is more like a shipping issue with the supplier, if it's owned wholly by them.
Are you saying the dough is fine but 1% of the batch wasn't cooked? Or rather, that 1% of batches isn't cooked, taking your entire downstream activity down until a new batch arrives?
Their engineers almost certainly are not the ones making the decision to let all the new traffic degrade their service for their existing paid customers.
That seems a bit crass but the underlying sentiment stands. Microsoft has more money than God. 4T valuation. When you say you’re worth that much, no excuses. Figure it out.
So your proposal is for Microsoft to invest their free cash in helping support a bunch of AI coders maintain their pet projects? Sounds like you should be a CEO!
There isn’t mission critical work being done on GH. You start charging more and the people in charge of finance at companies paying for GH will actually start paying attention. Those finance people don’t care rn because GH is affordable and helps get stuff done. There’s a threshold that exist at every company and once it’s crossed the people who can make money decisions start asking questions
They could throttle git operations to a level which isn't noticeable by normal 'human' usage but seriously slows down automated usage. These are basically DDOS attacks.
"sympathy" for one of the worst corporations in history is some insane corp bootlicking logic.
This is a multi billion dollar corporation. History of robbing and stealing from others of their labor or IP. History of enshittifying once great services.
Don't really understand why people are saying this is terrible advice. If you're 17, you should be learning everything. At that age, your brain is a sponge and your energy levels are the highest they will ever be.
None of us know what the future of work, education, or AI is going to look like. But your best bet is to become a life-long learner. Be it LLMs, musical instruments, physics, or business.
The app doesn't know. It just produces a widget tree, and the OS-provided renderer renders it this way or that way. In particular, it chooses not to render controls marked as "security-sensitive" when the rendering is intended for a screenshot; it could instead put empty boxes in their place, etc. The app has no idea and no control, AFAIK.
Open the amazon app. Take a screenshot. See a toast message informing you that the amazon app has detected you've taken a screenshot. Fucking used for fucking profiling.
I think at least in snap it makes sense, the communications are suppose to be ethereal and disappear after reading, I like to know when someone screenshots something and its no longer ethereal
There's an unspecified short/random delay between the user taking the screenshot and the notification firing. The trick used by Bluesky, Telegram etc is using a different technique. They (mis)use secure input fields to show an overlay when the user takes a screenshot.
Snapchat advertises that they detect screenshots. They try. But they're well aware that they can't actually know.
For example, their bug bounty program policy helpfully informs you that "screenshot detection avoidance" is not considered a vulnerability: https://hackerone.com/snapchat . That's because it's always possible.
I wish I could walk around NYC without airpods in. Someone mentioned the lunatics and others blasting music but the main issue honking.
The honking in NYC is unbearable. Add to that the noise that the subways make and the constant construction and you're risking actual hearing damage if you don't wear some noise cancelling headphones.
Can anyone comment on the inline AI tab completion performance of Zed compared to Cursor? Hoping to move away from Cursor as it seems like they break my settings every time they push an update and Vim motions stop working at random.
The only thing stopping me from leaving Cursor is their tab completion, which is honestly just incredible.
It's okay -- usable but definitely a downgrade from Cursor. Basic pattern matching completion works pretty well, but you won't be zipping around the codebase in the same way as with Cursor, which always seems to be thinking a few steps ahead.
The elephant in the room is that there's nothing quite like Cursor Tab. Copilot, Supermaven, Codecompanion, etc. don't even come close. As much as I want to use Neovim full time, I just can't walk away from Cursor Tab. I can live without Cursor Agent since I can just use Claude Code when I need an agent.
Until something comparable for Neovim comes out, I just don't see how I can switch back. I would happily pay for this. I'm sure there are a lot of people in the same boat as me.
I switched from Cursor to Neovim and at first I hated Neovim’s poor AI ghost text. Now I kind of like the break from heavy AI when I write code.
The vast vast majority of my code is written by Claude Code, so I don’t write much. But when I do it’s like an exercise in warding off programming atrophy
Most dont want that slop in their code. I have no issues with just writing the code myself. The "typing code" was never a bottleneck, and AI wont help for the real bottlenecks you have in an business setting.
Maybe someone can clarify this but I was also pretty appalled by the grammar in the Epstein emails until someone pointed out it could be an artifact of OCR or decoding issues.
Not sure why they would have to do OCR on emails. Were they printed out? On PDF for some reason? The decoding thing I kinda get but that you can easily point out because of all the equal signs.
reply