I just took a look at your game and I love how you can purchase a DRM free version of the game for Linux! That said, have you considered putting your game on GOG? (I personally prefer that platform).
> Upgrading to GitLab 19.3.6 on October 22 didn’t fix the vulnerability. The fix isn’t in GitLab’s own code but in a patch to Ruby, which GitLab applies only in its official packages. Our image used standard Ruby, so it didn’t get the patch.
> The bug is in Ruby’s regular expression engine, so any program that compiles regular expressions from untrusted input is exposed, not only GitLab.
I feel like this era of constant exploits/patches is really bad for those who build from source or make their own images vs those who rely on provided images and upgrade them often. I really can't think of a decent solution for those people other than being plugged in 24/7 to the latest news which is really unfortunate and must be exhausting...
At this point if you are self hosting you should have everything behind a VPN. Only companies with their own security and development teams should be putting things on the public internet.
Gosh I'm at a big company and I'm pretty sure all the self-hosted tools are only accessible through VPN-like tech anyway. The only truly public stuff is Jira, Confluence (which I assume it more to do with Atlassian's pricing of self-hosted) and Sharepoint.
Atlassian products (Jira Server) are no longer supported, and their more Enterprise offering dies in 2029, so everything is ending up in their cloud now yeah.
I know a lot of companies vibe-coding alternatives. I'm not an expert but it's perhaps even illegal for most European companies to migrate any customer data to Atlassian servers.
If someone wants to vibe code an alternative to a SaaS tool, my gut feeling is "no, this is bad and it will blow up" as vibe-coded projects tend to do.
But then the alternative is often paying $X million per year for licenses to a half-assed program that hasn't had a UI update since 2005.
I just don't have a good argument on why they shouldn't when the outcome will be just as dogshit, only cheaper and flashier.
They just started updating the UI and I am pissed off.
I don’t want them to change UI. I have plenty of work to do and have plenty of other tools that I use daily.
I have my workflows, I am swamped with work and now I have to change my workflows because someone changed options in UI where ones that O used are not there anymore.
I think there's ways to have a solid app by giving ai precise instruction about the architecture it has to follow. Especially by configuring read-only agent for security and quality and builder agents for everything else. It isn't perfect and I wouldn't do critical apps for a company with it but it's solid enough to replace a SaaS tool in my opinion. I think the real subject is wether this vibe-coder knows he'll have to maintain his app...
I'd pay embarrassing amounts of money for applications which do not get biweekly UI overhauls.
I'm still using the 2024 version of JetBrains IDEs because the UI hadn't changed in over ten years. I quit paying JetBrains when they ripped out the perfectly functional UI and replaced it with a VSCode reskin.
Please, for the love of god, give me software with a consistent UI.
The argument I'd make isn't about UI or features, it's about who owns security afterwards. The vendor has people whose whole job is patching things like this Ruby regex issue. The internal vibe-coded Jira usually has nobody, and the bugs that bite are boring ones, like a project page that loads for anyone who guesses the ID.
If they go ahead anyway, the ask is just that someone actually owns it after launch, not only builds it.
This person is just posting absolutely baseless claims, as if Atlassian, a company that basically exists purely on the back of B2B SaaS sales, has not considered EU data residency requirements in their core product offering: https://www.atlassian.com/software/data-residency
FWIW nearly every SaaS or IaaS is making bold claims about supporting data sovereignty, but unless it has been challenged in court I wouldn’t put much stock into it.
The fact that Jira/Confluence are being processed by their LLMs puts them in direct legal conflict with several statutes in the EU, its a matter of time for them to get hit by this and I’m betting that they think the fine will be lower than the revenue they acquire for doing it- and they’re hoping they don’t get hit at all.
This seems to be the common case for a lot of the non-EU tech companies, and it seems to be working.
Oracle is currently the market leader in “sovereign” compute for EU governments for example, as ironic as that is.
> This person is just posting absolutely baseless claims
You don't need to be so hostile. I know about the EU requirements but many EU companies have sensitive data on hosted JIRA / Confluence that simply cannot be in the possession of a non-European company. I know because I'm working for one. I cannot even remote-work from outside Europe. Some stuff cannot even leave the VPN. There's also the CLOUD act.
Yeah I thought Jira at least was all built on AWS these days. I've never set up a Jira Cloud instance but I'd be surprised if you don't chose where your data lives.
In short gdpr compliant means you can configure it but it must not be compliant by default and also there is no guarantee.
They are constantly changing their software and settings. More than once they believed that something is legal but can be risky for the atlassian customer.
For example training AI on metadata which they believe are anonymized but are maybe not enough changed.
Training data or metadata permisson can not be configured on all subscriptions. But as a data processor they must follow the contract otherwise they become data controller.
Also there are risks associated with privacy data framework, cloud act etc. to be non compliant just by data transfers to other countries.
Atlassian is a US-parented company and this means that CLOUD act can affect it. Even this only can make it unsuitable. Then there are questions to data-in-transit. Also nothing is legally tested. Vibe-coding an alternative is much safer if you have sensitive customer data (even simple support tickets).
For personal services, even something as simple as HTTP basic auth on the webserver dramatically reduces the attack surface by blocking all probers and drive-by attacks.
> We have not requested CVE identifiers in Ruby for these fixes, at least for now.
> Ruby has generally considered it unsafe to compile and use a regular expression created from untrusted input. We stated this in the advisory for CVE-2022-28738, where we nevertheless treated a double free in Regexp compilation as a vulnerability after a comprehensive assessment. We did the same for CVE-2024-27282.
It's a Bad Idea (TM) to compile a regular expression from an untrusted input.
Despite what anyone would try to sell you, a regular expression is a programming language. It's Always A Bad Idea (TM) to compile untrusted input into any kind of programming language. Don't string concatenate SQL queries, eval is evil, it's all the same idea.
A regex might be a programming language, a proper regular expression (which there are libraries for, like Google RE2) is just a regular language and I don't think you can even charitably call that a programming language.
Despite this Ruby should have requested CVE numbers. They created a patch that was available to install and prevented this attack. CVE numbers would have triggered automations to update Ruby in everyones systems.
We don't do rolling upgrades on steam boilers nor do we do it on our door locks. Nothing is perfect, we just need good enough and we are currently far away from good at all.
When it comes to software it hasn't always been the case. Before the Internet software was considered something stable and even now there are areas, like functional safety critical software, where this is true to some degree.
It's not that it can't be done.
It used to be expensive to a certain extent it will stay expensive, but modern languages, tooling and formal methods will hopefully mitigate that enough.
Most door locks are easy to pick btw. They're not protecting you with technical excellence, they're protecting you by making the thief spend a little time so the neighbours might see him.
What’s interesting with steam boilers and other types of OSHA related injuries is that it’s not enough that the person who’s doing the stupid thing dies: you have to enforce it even more and even higher. People won’t value their life.
> I really can't think of a decent solution for those people other than being plugged in 24/7 to the latest news which is really unfortunate and must be exhausting
The future is near constant burning of tokens in your software factory to do this automatically. The labs create the increase in demand for their own product.
Turns out building the world on an infrastructure of lies has finally caught up with us.
The way LLMs are developed now is but one possible future we could have experienced. Imagine the future where the LLMs were built by a few governments with black budgets and hidden until a war or some other large event broke out. Then suddenly most of the computers and digital infrastructure shits itself as everything goes wonky.
At least this way we get to see that a shared global infrastructure built on the idea that security is someone else's problem was a really bad idea and we might have enough time to fix it or pull enough plugs we don't go back to the dark age.
That's a for now problem. The future is probably less software surface area exposed to the internet, the removal of unsafe languages from software exposed to the network, services like Cloudflare deploying firewalls to detect exploit attempts, and services that can keep a live database of currently active bots and residential proxies to IP block them.
Alongside the current scanning and patching of software.
I don't disagree. Also probably more logs and more metrics since LLMs can scale instrumentation, signals and auditing far beyond what humans could ever process, which in turn, more tokens. Probably larger Datadog bills too.
The world has changed but 15 years ago we thought it was utterly insane to just pull whatever $latest from $internetsite.probablytrusted.org as part of your build and deployment process.
The patch for Ruby came out 6 days ago. Most Rails apps do not use the Ruby from the package manager but have a hardcoded version in the repo. So unless you have developers releasing a PR to update ruby the moment it comes out, you aren't safe.
This might feel like a "Yeah obviously" moment, but the fact is that very few Rails apps are running the latest ruby release the moment it comes out.
> build from source or make their own images vs those who rely on provided images and upgrade them often.
I mean - this incident brings it into sharper focus - but it's always been thus.
As a Go developer I often tell people "Be wary of vendoring, you are creating a cache of dependencies that need to be managed and maintained"
Or people that "pin" dependencies, they face the same difficulty.
Why not just use "official" (in this case) images - sometimes they lag too - hell, most Linux distributions lag so much with Go compiler packages that, again, I routinely tell people to avoid them in favour of watching the Go official page
In go, vendoring dependencies just copies the deps out of the module system into the local tree and it's managed the same as not vendoring. It won't even build of the go.mod has changed without also updating the vendor tree.
Right - a copy is a cache, and you have to manage that; when to update it, when vulnerabilities are discovered in the deps you have copied, it's up to you to change the version in your cache.
While I agree with your statement overall, as someone who drives everyday in LA alongside Waymos without anyone inside of them I must disagree with this aspect of it.
A better formulation would be: we don't know if we have self-driving cars.
Because if it became fashionable for pedestrians to wear large green dresses with red balls, and everybody would wear that, we have no idea what the AI would do.
This is a general point rather than an argument supporting the GP comment, but it can be shockingly hard to know what the correct response to particular situations is, or predict how a system will respond. For example, one system I know had difficulty differentiating people laying on the road from unmapped speedbumps (particularly at night), but was very good at recognizing them standing up. Someone spent a lot of time testing with dummies until perception improved. Another had no idea how to handle people in fursuits and halloween costumes, because it was incorrectly applying animal behavioral models to pedestrians.
That's why the latest self-driving systems include reasoning models that draw from the massive amount of knowledge embedded in LLMs. Language evolved to communicate things that are not here and now after all.
We still have no general self driving cars. What we have is a bunch of geofenced tests. While we are closer than ever, we are still seemingly a long way away from truck drivers being obsolete was foretold over 10 years ago now.
I don’t know much about this so I’m curious. Is the reason why GPUs aren’t preferable for large scenes due to their limited VRAM or a different aspect of them?
Not in a serious manner but I created a testing harness for a Nintendo 3DS game I'm making that uses the OpenAI Decisions API. The main advantage is the speed (~2-300ms per input) which I really need for this purpose.
I just found out about this an hour ago while browsing the Pi docs. I'm really excited to use this as an alternative to the herdr pi extension as I prefer minimizing the number of extensions I have. Also, this could be great to use with a CI tracker (though I'd need one that works with Forgejo).
As someone who mails a bunch of letters (as a hobby) I'm surprised to hear stamps being used this much nowadays! Just out of curiosity, how often did inmates actually use them to mail letters, or was it predominantly used as a form of currency?
Great question: I would guess about 15% of the inmate population sends mail home on any given day. That number certainly jumps up if a facility is locked down, probably by at least double.
I have a friend who did some time a decade or so ago (I'm uncertain if it was a federal or state prison) and one thing I've noted about him is how much he values books, even stuff like manuals and outdated reference material (we're in a volunteer group and we were cleaning up a room with a bunch of trash). His explanation to me was that in prison there wasn't much to do and so he spent his time reading as much stuff as he could get his hands on.
From your experience, would you say that the average rate of reading amongst inmates is higher than the general population? In person I've hardly met anyone who reads books other than for school assignments.
> In person I've hardly met anyone who reads books other than for school assignments.
I know a ton of people in person who read a lot for pleasure, but I'm sure it's kind of self selecting. Are you in college? College was the time in my life I read the least.
Also, while I haven't been to prison, I've been to jail for a short cannabis related stay in the early 2010s. There is just so much less to do there that I ended up reading a whole book (2-300p) every day I was there. Many others had a book, but I'm not sure they were actually reading them. A lot of people in there were reading the Bible or some religious literature. Anyway, just my anecdote.
Nah, I'm in my 20s but already graduated. That said, most of my friends are guys from a church group and are 40+. I take back what I said a bit though. I know two of them who like to read spiritual stuff, but the rest don't read books and do social media or TikTok only.
I see no reason why they can’t target both at the same time. Personally I know two people who never would’ve paid for ChatGPT a year ago but now do so. I think for the average person it’s going to be a slow and difficult conversion process to paying customers.
reply