This is bad.
Uncategorized
149
Posts
33
Posters
0
Views
-
@xgranade why do you consider open-slopware a mistake, btw?
@outfrost I don't, per se, but I consider the way I participated in it to be a mistake, and one that got people hurt. Any time you make a list of people, no matter your intentions, that takes caution --- caution that I did not personally put into practice. I don't get to hide behind my intentions on that.
-
@outfrost I don't, per se, but I consider the way I participated in it to be a mistake, and one that got people hurt. Any time you make a list of people, no matter your intentions, that takes caution --- caution that I did not personally put into practice. I don't get to hide behind my intentions on that.
@xgranade gotcha, so it's not about the main list itself, but callouts on specific maintainers?
-
@xgranade I also dislike it, but the cat's out of the bag, even if it wasn't allowed people would still be using it, just without revealing it
@MissingClara @xgranade That's a bad argument against having a policy. Policy is a statement of who does and doesn't belong. If they're using it without revealing it, trying to launder code with fraudulent provenance into your project, that's highly malicious behavior worthy of a ban from the project once they're caught. And it's a signal to your good contributors that the project is healthy and not going to be turned into irreparable garbage by slop bros.
-
@xgranade gotcha, so it's not about the main list itself, but callouts on specific maintainers?
@outfrost It's complicated. There's very valid critiques of the list, and also bad faith misrepresentations of what the list was. I can only speak to my own actions, given how complex everything got.
-
@iampytest1 @theorangetheme The idea of putting a noreply email address on your commits is extremely funny to me. What exactly is the point of putting an email address on the commit message at all, then? It's not supposed to be your ID badge. What do you think is the reason that standard was created?
-
@outfrost It's complicated. There's very valid critiques of the list, and also bad faith misrepresentations of what the list was. I can only speak to my own actions, given how complex everything got.
@outfrost I put more of those thoughts together when I first apologized, and when I also tried to be clear about what parts I did not and still do not apologize for.
-
@iampytest1 @theorangetheme The idea of putting a noreply email address on your commits is extremely funny to me. What exactly is the point of putting an email address on the commit message at all, then? It's not supposed to be your ID badge. What do you think is the reason that standard was created?
@iampytest1 @theorangetheme Just another example of how LLMs degrade code quality. If a commit is "authored" by Claude, there is absolutely no one accountable to that code. If you want to reach out to the committer, it goes to an unmonitored email address. Great! Very healthy for our systems.
-
@joXn I'm not here to critique specific individuals, and I'll ask that you don't use my replies to do so either. In particular, this is a large systemic problem across OSS, and while I can think of a few specific bad actors making things worse on purpose, I don't think that's the most common modality by far.
Besides, if having worked for msft is QN automatic disqualification, I'll disclose that I worked there about 5.5 years.
-
@iampytest1 @theorangetheme Just another example of how LLMs degrade code quality. If a commit is "authored" by Claude, there is absolutely no one accountable to that code. If you want to reach out to the committer, it goes to an unmonitored email address. Great! Very healthy for our systems.
@iampytest1 @theorangetheme The only purpose behind it is anthropomorphizing Claude Code in order to sell "AI". You wouldn't put Visual Studio as a co-author on your commit. But because Claude is supposed to be "a person" (read: slave), we pretend this tool is an equal author, even though they don't exist and it's impossible to contact them.
-
@ancoghlan I mean, yes, but also AI is DDoSing the heck out of that process? More that point I was getting at is that I wasn't able to find any policies against AI-generated code in the first place, so there's very little in the way of safeguards to prevent more such commits in the future.
@xgranade @ancoghlan I don't know that any denial of service has *yet* taken place in CPython, but it's definitely happening in OSS writ large.
I'm thinking again of Naomi Ceder's excellent PyCon US talk about the gift economies of open source and her words of caution to guard against financialized interests… I don't know how to square these contributions in that perspective beyond the amount of personal trust invested in people with the commit bit.
I respect and trust Serhiy and Gregory quite a bit! If such a commit were authored by a core developer whose work I am less personally familiar with, I might feel very differently…
-
@jo My thread detailing what I do and do not apologize for may be a good start?
-
@xgranade @ancoghlan I don't know that any denial of service has *yet* taken place in CPython, but it's definitely happening in OSS writ large.
I'm thinking again of Naomi Ceder's excellent PyCon US talk about the gift economies of open source and her words of caution to guard against financialized interests… I don't know how to square these contributions in that perspective beyond the amount of personal trust invested in people with the commit bit.
I respect and trust Serhiy and Gregory quite a bit! If such a commit were authored by a core developer whose work I am less personally familiar with, I might feel very differently…
@SnoopJ @xgranade @ancoghlan Examples of this happening for reference purposes: https://www.pcgamer.com/software/platforms/open-source-game-engine-godot-is-drowning-in-ai-slop-code-contributions-i-dont-know-how-long-we-can-keep-it-up/ https://arxiv.org/abs/2601.15494
-
@outfrost I put more of those thoughts together when I first apologized, and when I also tried to be clear about what parts I did not and still do not apologize for.
@xgranade i see, thanks a lot for the context!
-
@xgranade @ancoghlan I don't know that any denial of service has *yet* taken place in CPython, but it's definitely happening in OSS writ large.
I'm thinking again of Naomi Ceder's excellent PyCon US talk about the gift economies of open source and her words of caution to guard against financialized interests… I don't know how to square these contributions in that perspective beyond the amount of personal trust invested in people with the commit bit.
I respect and trust Serhiy and Gregory quite a bit! If such a commit were authored by a core developer whose work I am less personally familiar with, I might feel very differently…
@xgranade @ancoghlan I definitely cannot see how the use of the tool could be scaled up from where it is *right now* to "a substantial fraction of contributions use it" without burning up more human time.
I hope that the claim of ownership signified by the CLA is taken seriously in the inevitable-under-scale case of someone who contributes code they *don't* own, but I guess it's easy to wander into hypotheticals here.
I can say for sure it is a net negative in my view of the project and will probably say so in the next PSF survey, but I don't know how active I'm going to be about that displeasure until then. Guess it depends on what else changes.
-
Part of the problem with doing so is.... well, now what? It's not like a Python project can just... stop being a Python project?
But I think it's important to at least understand the scope of the problem.
@xgranade doesn’t micropython also run on Unix?
If you don’t depend on all those C extensions, maybe that will do…?
-
@xgranade doesn’t micropython also run on Unix?
If you don’t depend on all those C extensions, maybe that will do…?
@xgranade (I mean, CPython now also requires Rust 😾 so… that would be wise anyway.)
-
As an addendum, I'm using Python as an example here because it's near and dear to my heart. This is not "Python in particular is exceptionally bad," this is "a very bad thing has been happening in OSS *in general* and Python is now in that blast radius, which makes it harder for me to personally ignore."
@xgranade worse, Linux (the kernel) pretty much is…
-
@joXn I'm not here to critique specific individuals, and I'll ask that you don't use my replies to do so either. In particular, this is a large systemic problem across OSS, and while I can think of a few specific bad actors making things worse on purpose, I don't think that's the most common modality by far.
Besides, if having worked for msft is QN automatic disqualification, I'll disclose that I worked there about 5.5 years.
@xgranade Thank you for clearly expressing a boundary here, which I will be glad to respect.
My point was not that “working for Microsoft” is disqualifying — I also worked there for about that long (on a Python-related product, too!).
It goes more to a broader question of what puts any OSS project squarely in the LLM “blast radius”—when a significant set of senior contributors really believe in AI as a “useful tool”, that will be more “useful” with “improvement”, that’s probably a problem.
-
@ancoghlan I mean, yes, but also AI is DDoSing the heck out of that process? More that point I was getting at is that I wasn't able to find any policies against AI-generated code in the first place, so there's very little in the way of safeguards to prevent more such commits in the future.
@ancoghlan @xgranade not only AI people are DDoSing that process, the process wasn't intended to review plausible-looking extrusion of github paste from your own tools. The threat is new, increasingly documented as a cognitive hazard and it's scary to see people with commit access to the backbone of the ecosystem just YOLO it with hubris.
-
@xgranade Thank you for clearly expressing a boundary here, which I will be glad to respect.
My point was not that “working for Microsoft” is disqualifying — I also worked there for about that long (on a Python-related product, too!).
It goes more to a broader question of what puts any OSS project squarely in the LLM “blast radius”—when a significant set of senior contributors really believe in AI as a “useful tool”, that will be more “useful” with “improvement”, that’s probably a problem.
@joXn Thanks, I really appreciate it. Anyway, yeah, I agree that that's a problem when there's not active opposition from people with the ability to influence the direction of a project. That's a very broad problem in OSS right now, though... I'm just sad that it's hit as close to home.
Feed RSS
Gli ultimi otto messaggi ricevuti dalla Federazione
-
Can my English readers guess for what kind of article I drew this diagram?
-
(Questa è difficile)
-
Generazione Italia
-
Com'è, troppo colorato?
-
Reggio emilia: al bosco ospizio saranno abbattute il 70% delle piante. un anno dopo prosegue la mobilitazione
@anarchia
Un anno fa, il 25 febbraio 2025, nove cittadini bloccavano i lavori di sfalcio preliminari all’intervento di Conad nell’area dell’ex Casa di riposo a Reggio Emilia, venendo poi denunciati per violenza privata,
-
@Gina @bert_hubert ik zucht met je mee. En zo sneu dat daar ook gewoon hele goeie developers, engineers en architecten rondlopen die volledig gepasseerd worden in het hele verhaal, maar zich waarschijnlijk wel kapot schamen.
-
@bert_hubert ik heb zelden iemand zo graag gevolgd wiens toots me toch geregeld zo verdrietig maken
-
America, and probably the world, stands on a precipice
Link: https://garymarcus.substack.com/p/america-and-probably-the-world-stands
Discussion: https://news.ycombinator.com/item?id=47171771
