The LLM part is bad news but it isn't the sign.
-
@mhoye this is a great thread, and I agree with all of it except this tangential point.
Matrix moderation tools are a chaotic mess. Some of that is inherent distsys problems, some of that is missing / bad protocol design or poor integration. Not all of it is directly related to moderation being a reactive afterthought, but is compounded by the combination of that with general matrix desync artifacts.
Speaking from experience doing matrix moderation, notably during an especially nasty period a year or two ago, there is plenty of capability left to build. While the more centralised platforms have some inherent advantages here, that's not the only reason their moderation capabilities are easier to operate effectively.
There are tools. Many different variants, even. That they are better than what IRC has is not in dispute, but that's the lowest of low bars.
ETA: The people building the tools are working on the fringes and get easily burned out, for obvious additional reasons beyond all the normal and accelerating reasons everyone else is These Days.
@uep so, it’s been 3+ years admittedly but that was not my experience at all. I don’t want to tell you you’re holding it wrong, and I haven’t had an opportunity to flex the new Draupnir tooling, but my experience was definitely not what you’re describing.
(Slack’s moderation tools basically don’t exist and IRC’s are trash, on this we agree)
-
@uep so, it’s been 3+ years admittedly but that was not my experience at all. I don’t want to tell you you’re holding it wrong, and I haven’t had an opportunity to flex the new Draupnir tooling, but my experience was definitely not what you’re describing.
(Slack’s moderation tools basically don’t exist and IRC’s are trash, on this we agree)
@mhoye yeah, there's good stuff, but ultimately there were two key issues (and lots of smaller ones and compounding effects)
it's largely based on reactive bots acting to delete things after the fact, so The Stuff still gets seen, and seen differently by people on different servers, clients, etc.
the attackers know the limitations and exploit them, moving to parts of the system that are not covered. Rather than spamming channels, they send DM invites with profile pics, for example. Such arms races are inevitable, but matrix can't adapt across the ecosystem nearly as quickly.
There are good community tools. Matrix itself gets used as transport so that one community's banlists get published to a channel for others to apply.
Again, though, I'm only reacting to the "nothing comes close for mod tools" aspect. There's not much else around like matrix, for good and bad. And none of it detracts from the rest of the thread.
-
@denisbloodnok @waffles irc.mozilla.org was a public server, yes.
@mhoye glad to hear that it wasn’t internal abuse. but damn, would not want a job where i was expected to listen to that.
-
@gsnedders That didn't happen at the same time; the reality is that IRC usability was, compared to what was available, just so bad for so long that long before a single Mozilla slack instance was blessed, long before I was sent to get us an alternative to IRC, there were already _dozens and dozens_ of private shadow Slack instances all over the place for various teams and projects. Like, on the order of 180 of them. One of the biggest impetuses for moving to Slack was... most people already had.
@mhoye My recollection is very much that enough people were vaguely on IRC, even if a lot of the discussion was happening in the Slacks, to an extent that is no longer true with Matrix?
Like, you had both having to change the communication channel that was infrequently used by many people, plus the single Slack instance made possible more cross-team communication to happen within the private channels.
Certainly the stuff I was around (mostly platform, and especially a bunch of the tools & automation stuff), it was very much around the time of the Matrix move that much more discussion moved into private spheres.
(And like, I’m not gonna say that other corporate driven projects are better in this regard, but forcing people to transition to a new tool can definitely make this worse. And I’m certainly not defending IRC here, but it was at the time of the migration that things got worse as an occasional contributor.)
-
To explain what I mean by this: for a long time the Perl community took great pride in how difficult and terse the language was. The old joke that well-written perl was indistinguishable from line noise wasn't just a joke, it was a badge of honour.
That attitude was sustainable when the internet was small and alternatives were scarce. But the moment tools appeared that prioritized things like human ergonomics (python) and human interaction (ruby) the perl community started aging out and dying.
@mhoye That's not quite what happened with perl, but what actually happened was a lot closer to the narrative from the rest of your thread so it doesn't detract from your point at all. Part of it was also that Perl 6 sucked all the enthusiasm away from trying to do anything to improve Perl 5 which was hard for both technical and social reasons.
-
Anyway: The worst thing an organization's leaders tolerate is what that organization becomes, and the blast radius of a single toxic contributor is huge and largely invisible. For everyone who speaks up to say this behavior isn't acceptable, there are dozens or hundreds of smart, competent people who peer into your community, see that this is how you live and move along without your ever knowing they exist, on their way to a better life somewhere else.
We don't have to live like this.
@mhoye i work and plead mors on the build tool and compiler and rest of tech tool aspect side of things with the same arguments.
And I think it is far less important than the one you point out (and probably harder to fix funnily enough).
So thank you for expressing it
-
RE: https://woof.tech/@techokami/117439226690379216
The LLM part is bad news but it isn't the sign. It's the talking about planning succession while simultaneously complaining that nobody new wants to learn, that's the sign.
Y'all the kernel is in deep, deep trouble. We saw what happened to Perl and the kernel is in deep trouble.
@mhoye What if the two are linked, and the succession plan /is/ to drive upstream off a cliff in order to force a hard fork.
-
-
Today the number of people under 50 who'll reach for perl or raku or who cares what it's called as their first choice of tool for anything new is zero. In practice Perl is an esolang now.
Torvalds' position is that contributing to the kernel should be _more_ difficult for everyone else, but, through LLM usage, easier for existing kernel developers.
Ted Tso's adamant position that "you can't make us learn Rust" is the same insularity.
If the kernel was a country, we'd recognize this.
@mhoye As a 34 years old Perl dev, I have to disagree on the number being zero.
-
RE: https://woof.tech/@techokami/117439226690379216
The LLM part is bad news but it isn't the sign. It's the talking about planning succession while simultaneously complaining that nobody new wants to learn, that's the sign.
Y'all the kernel is in deep, deep trouble. We saw what happened to Perl and the kernel is in deep trouble.
@mhoye I think they didn't complain about the lack of people willing to learn though. There is a deluge of patches from people who just want to get them in with LLM help without understanding how any of it works. But there are also good faith contributors
-
Today the number of people under 50 who'll reach for perl or raku or who cares what it's called as their first choice of tool for anything new is zero. In practice Perl is an esolang now.
Torvalds' position is that contributing to the kernel should be _more_ difficult for everyone else, but, through LLM usage, easier for existing kernel developers.
Ted Tso's adamant position that "you can't make us learn Rust" is the same insularity.
If the kernel was a country, we'd recognize this.
@mhoye I'm over 50 and I never reach for perl. Haven't for a decade.
-
And for Linux as a plausibly community-managed meaningfully open as in open project, that clock is definitely ticking.
The end of this road is some corporate-funded retirement home for understaffed and underfunded projects once enough of the current devs have aged out without anyone growing into those roles, with private forks at companies that relied on the kernel, also privately relieved there's no need or reason to deal with antediluvian developer processes or bother giving back anymore.
@mhoye The "just stop onboarding new maintainers" thing was even more shocking to me than the heavy AI psychosis, although they're probably related.
-
@lykso@tiny.tilde.website @hipsterelectron@circumstances.run @mhoye@cosocial.ca @froztbyte@mastodon.social In my case I have no belief my license works as more than a statement of intent/principles.
So I try instead to compensate in things which I believe the generally applicable contribution to autonomy remains applicable regardless of whatever corposcum might copy it (whatever bastardized copy may not retain that, but the one I will continue to provide will).
I believe finding the source of things matters and that the tooling for current code helps a lot.
Ultimately of course even with protections, corposcum have very well demonstrated in book publishing just how effectively they can drown out one’s signal so that one is never heard of anyway (copyright was a monopolist and censorship tool first & foremost and it is what it does best; the same is applicable to patents in a very similar way). The system has always protected the monopolists and corposcum first, anyone else being a pretense or incidental at best.
@lispi314
By the way of my Anti AI license project I read some court decisions and this was very disappointing for me. At least in Germany it is nearly impossible to run a (national/international) project of distributed volunteers which share code in the open. There are very weak protections for, say, GPL to be obeyed. In other countries it is probably the same.
I also believe I now understand what these EULAs are supposed to achieve. Funnily enough they seem to be non-binding in Germany if not printed outside the CD you buy. -
To explain what I mean by this: for a long time the Perl community took great pride in how difficult and terse the language was. The old joke that well-written perl was indistinguishable from line noise wasn't just a joke, it was a badge of honour.
That attitude was sustainable when the internet was small and alternatives were scarce. But the moment tools appeared that prioritized things like human ergonomics (python) and human interaction (ruby) the perl community started aging out and dying.
@mhoye I tend to disagree a bit about perl. Yes, it looks like line noise to the uninitiated, but the amount of initiation is not huge. I understand that this is pushing away beginners who in Python can just start typing code. OTOH, perl was invented by a linguist and it is intended to be dumped into the Wernicke area and parsed there, whereas python looks like line noise piped thru `gzip -d` , resulting in phrases exceeding the input buffer size of said area. Disclaimer: I'm 64 and use raku.
-
@sackofbones @Netzblockierer @torvalds @mhoye Perhaps consider trusting a project (and its leader) that have delivered increasingly capable kernels and a strong development community for 35 years?
I am not thrilled (to put it lightly) about this rush to LLMs, but I don't think that this sort of attitude makes things better. -
@Lunaphied @fedward @mhoye Yes, raku is fun and good. It is essentially perl with the bad things (ambiguous syntax, lack of static typing and good OO) fixed and cool stuff thrown in (grammars, hyperoperators,...). It feels like complex analysis versus real analysis in math: perl5 and real functions can be a nightmare full of traps and counterexamples; with perl6 and complex functions almost anything that looks plausible is proven to be correct.
-
@paulmckrcu oh, on a personal level I think I've been round that loop a few times. A lot of what U. G. Krishnamurti said to people who thought they wanted to replicate his path has come back to me over the years!
@srtcd424 Fair enough!
-
@mhoye tldr; #LLMs are here to stay. People who actually deliver working solutions that shape the world have spoken. Measures to limit slop are being developed. #Linux becomes even more commercial.
I personally hope more open models will arrive, GPUs will become cheaper and universal income will help people who wont be able to find job in such toxic and draining field as computer science.2026 welcome in ultimate hell everyone! Dont forget to enjoy the ride!
@grayrattus @mhoye How do cheaper gpus make universal income possible? The wealth hoarding class can make universal income a reality right now.
-
@grayrattus @mhoye How do cheaper gpus make universal income possible? The wealth hoarding class can make universal income a reality right now.
@Howitzer105mm @grayrattus You can see it's definitely a well-thought out position.
-
@Howitzer105mm @grayrattus You can see it's definitely a well-thought out position.
@mhoye @Howitzer105mm I wrote that I hope for 3 things: cheaper GPUs, more open models, universal income.
Without those next 5 years will be tough as many IT engineers will land on the streets with huge debts and burnouts.
