oh wow you’re just like this all the time huh

no wonder you came in here to scream for a disgusting chicken sandwich incorrect one of my posters about their use of a common English phrase and post yet more LLM apologia barely disguised as critique

yeah nah we don’t need this centrist AI booster crap here but thanks anyway

But from all sides really, also wild to just claim they don’t know what a zero day is and that’s just made up.

some motherfuckers really see a security vendor claim a zero day can’t be exploited at scale for a local application, ignoring gigantic classes of vulnerability enabled by misconfiguration, combined exploits, or malware, and go “woof, maybe it’s true! they do make my favorite password manager after all, who are you to say they’re wrong” as a bunch of Russians walk off with their bank info

you like 80% of the claptrap keepassxc posts? no wonder you came into this kfc asking for a double down. we haven’t even served those since, like, the mid-2010s

the project’s sudden commitment to code review excellence is the exact same shit every other project pulls when there’s justified backlash in response to a policy that allows, and therefore encourages, slop code. that keepassxc keeps officially posting through it, defending code-oriented LLMs as “generally accurate”, and fucking up and showing that they don’t understand their own threat model, is the double down. I don’t particularly give a fuck that they’ve remained remarkably consistent in their policy of accepting garbage into their codebase, or that their blog’s response to the backlash has been, golly gosh, so measured! if this is how their team conceptualizes risks to a piece of software whose breach would constitute a catastrophic event.

“blackjack”? kfcs don’t allow gambling, what the fuck are you on about

And it might be debatable whether that’s a risky game.

debate the merits of slop code in a password manager elsewhere, thx

it’s only a double down if it’s a kfc sandwich where the bread is replaced by chicken. i see no chicken sandwich here, alleged posters, unlike in fuck ai where it’s chicken sandwiches all day

itt some fucker thinks slop code in a security-critical project is justifiable

froztbyte’s criticism crossed the line by a bit for a couple of admins who weighed in, and they’ve been warned to ease up. reporting a post like that isn’t bannable; we’ve got more context for a report like that than we do for some rando doing a drive-by report for a tone rule that doesn’t exist, for example.

blue misused the report system in a way that wasn’t accidental or incidental, and we felt the best course of action was a cooling off period. given that they’re welcome back in less than 4 days, I’d prefer to leave it at that.

the public modlog, linked from the instance sidebar:

Your claim sounds reasonable to me,

quiz show “wrong” buzzer, poster disappears down a trapdoor

Pay no attention to ChatGPT, we purposely trained him wrong… as a joke.

yeah ok

lightly used thinkpads are the classic choice for this — IT departments buy high spec ones then dump them for cheap a few years later in surplus sales or on eBay, and there are usually repair manuals and spare parts readily available. usually you can type the specific model and generation into a search and get a wiki page or at least a couple blog posts reporting how well they’re supported under linux, and Lenovo seems to intentionally do very well on compatibility since Linux compatibility is a nice checkbox for an enterprise laptop to have. just be careful you don’t get bamboozled into buying any of Lenovo’s consumer laptops, since they tend to be a fair bit cheaper and don’t have the same compatibility guarantees, repairability, or ample spare parts availability.

you know, I shouldn’t be surprised by the extremely toxic lead developer to prompt enthusiast pipeline, but… slopcode in gzDoom of all things? fucking why?

uggggh yep sorry, I meant to deploy some changes that would mitigate this a couple nights ago but had a bunch of things crop up. I’ll do my best to work it in tonight!

some specifics for the changes to expect:

• iocaine finally
• some better nginx settings to kill likely scraper connections faster

today in I fucking called it fedora aka mostly red hat has decided to allow slop code in a way that violates even their utterly mid stated principles around the tech

if you’re downstream from any fedora packages (and I don’t know the scope of this policy so it might be safe to consider anything owned by red hat in general to be tainted — yes I realize most of us are downstream from a bunch of red hat shit) it might be time to evaluate an alternative if available

of course the organization I know primarily for platforming fascists and astroturfing on YouTube was secretly an even worse grift and somehow tied in with Yarvin, why wouldn’t it be

given that Rossmann’s at the head of this thing too, I’m starting to regret not taking GrapheneOS (who, notably, were also a target for this grift) seriously when they said Rossmann’s involved in a bunch of terrible shit. the right to repair deserves a better figurehead.

sometimes crime pizza is fucking fantastic! they just let somebody’s grandpa cook and consider more expensive ingredients an investment in their cover. same idea as Capone’s soup kitchens but much easier to hide money with

well there you have it

bitter winter adult it is

nah sorry about that, the scrapers took the opportunity to knock us offline again so I did a little bit of impromptu maintenance to make us more rugged against the same type of failure in the future

the next work I do around this will be significantly more planned because it’ll be iocaine

I mean, search has always been built on some kind of LLM. That’s how you convert a query into a list of page-results.

no it fucking hasn’t. the stemming and page ranking algorithms used in traditional search have absolutely nothing to do with LLMs.

shit, neither stemming nor PageRank as originally defined even have a machine learning component. here’s postgres’ full text search suite, which literally converts a textual query into a list of results (sans page ranking, which is out of scope for a database) in a manner suitable for a production search engine, utterly without any machine learning or other stochastic crap.

what an unpleasant asshat