They make incorrect predictions of text to respond to prompts.
The neat thing about LLMs is they are very general models that can be used for lots of different things. The downside is they often make incorrect predictions, and what's worse, it isn't even very predictable to know when they make incorrect predictions.
I've even had to prove my salary when applying for apartments. No loan involved. The first time that happened I didn't have a w-2 yet, so they called my employer to check.
Overlooking your repeated assumptions of my incompetence at GitHub administration, neither of the 'why not just' steps you countered with will result in the banner that says 'This project is archived', which is a key component of my intentions. (For those who overlook it, the other characteristics of archival of course still suffice; I get all three for one UX interaction rather than three!)
Which of these projects is more likely, relative to the other in its pair, to be targeted by users whose expectations are presented disrespectfully by whatever means (let's assume email) the users can discover?
1a) A project that has recent commits, but has pull requests and issues disabled.
1b) That exact same project, with the banner "This repository was archived by the owner" shown on all pages and objects within it.
2a) A project that left a work in progress unfinished six months ago, but has pull requests and issues disabled.
2b) That exact same project, with the banner "This repository was archived by the owner [six months ago]" shown on all pages and objects within it.
As one can reasonably predict, archiving when I'm not actively pushing commits turns out to be an effective way to stem the tide of jerks who otherwise pick a fight by email/irc / discord/forum / blah/blah / etc., in hopes of persuading me to commit further resources to their unpaid benefit or to finish something I don't care to work on finishing right now / this week/month / quarter/semester / year/decade / ever. It's archived, so clearly it'll never be finished, which helps enforce an appropriate calibration of expectations upon those desiring the outcome — and if I someday finish it, hooray, but no one has any plausible way to justify any expectations to the contrary, no matter how hard they wish otherwise. Thus why I recommend using archiving to remove the social pressure component of working in public on GitHub.
Of course, if you have a better solution than you've offered so far here, I'm certainly willing to consider it.
You could add your own banner that makes it clear what the status and expectations of the project are. In fact, that is probably a valuable thing to do if the project is archived too.
If you archive it, then sure user's are less likely to send you disrespectful messages by whatever means (assuming they can find such means), but they are also less likely to use it, because you have basically said "this project is abandoned". Which if you don't want users at all, I guess that's fine, but then I'm curious why you bothered to publish it publicly at all.
The thing I’m publishing is currently the only public source of that information and code worldwide. For those who need it, there is literally no competition, unless someone else opens Ghidra and reproduce my efforts over several long years from scratch. I don’t publish because I want to play numbers-go-up with the world population; I publish because I invested years of research for my own benefit and wish those who come after me to have a starting place further along than where I had to begin. Sure, it’ll benefit users, but ‘on the shoulders of giants’ only works when you publish at all.
There are only maybe a few thousand people at most worldwide who could benefit from what I’m doing, and only 1% of them might actually go looking for this. Can you imagine working as an open source maintainer on a project with a hard cap on audience size of a hundred people? Seems fine to me; I’m one of those hundred, after all :)
> at the same time accusing others of historic conflicts of interest
Collabora clearly has a conflict of interest, as their Collabora Office products both benefit from, and compete with LibreOffice proper. They even allude to that conflict of interest in the next sentence:
> overriding past board and engineering steering committee decisions and violating their own processes to drag code out of the attic to enable competing with their largest single contributor
A non-profit dedicated to promoting open source software should do what is best for that project and its users regardless of if doing so steps on the toes of corporate sponsors.
"Collabora clearly has a conflict of interest, as their Collabora Office products both benefit from, and compete with LibreOffice proper."
But that was the point of TDF: being an umbrella for the community and the commercial partners. Commercial partner having its own interests is something that was expected and encouraged as long as the work would be done for the common good on the LibreOffice codebase. For the TDF itself there are COI policies in place as well from the very beginning.
"A non-profit dedicated to promoting open source software should do what is best for that project and its users regardless of if doing so steps on the toes of corporate sponsors."
That kind of thinking is what brought this result - especially when the corporate and the non-profit are supposed to be partners and work together for the common good. The result should be dialog to find a compromise that would work for both sides. So in this case forcefully reversing a previous decision and ignoring the de-atticisation rule (having active developers working on the code-base) was a needless aggressive move that just worsened the situation. And the reason for the move according to the TDF board was to "start the discussion".
Note that the proper procedure for de-atticisation of the LOOL codebase would be to confirm there is notable developer activity. Users wanting LOOL is NOT enough and the TDF board has clearly ignored one of its own rules.
This plausibly demonstrates why a nonprofit may not be a great vehicle for some free software projects - while the nonprofit should do whats best for the project, if the main work is done by commercial sponsors then it’s crucial those sponsors feel the relationship is beneficial.
The reality is free software office apps require significant professional development input. Apache Open Office is the obvious example.
It’s a classic version of the tragedy of the commons. If Collabora goes off to its own thing, I struggle to believe they will maintain the development rate with new devs, and without development the TDF sponsorship will fall off.
I hope we are not looking back in two years time regretting this.
> if the main work is done by commercial sponsors then it’s crucial those sponsors feel the relationship is beneficial
But if the sponsor is getting a tax write-off because of their donation to non-profit to do work on the project that they would have done anyways for their commercial product, then they are basically just using the non-profit to avoid taxes, and while I'm not a lawyer, it wouldn't surprise me if that is illegal, especially if the company also controls seats on the board of directors.
> a nonprofit may not be a great vehicle for some free software projects
I've frequently wondered if we need some new kind of structure for funding open source projects works kind of like a non-profit but is more lenient in some ways, like allowing some kinds of business transactions in addition to accepting donations, and maybe you don't get as much of a tax deduction for donating to it. I don't know exactly what that would look like though, and it would probably be difficult to get right.
You're considering open source development as just another commercial endeavor. The fact that this is done by a nonprofit organization means it's pursuing goals that are not strictly commercial, and that is fine. Think about the GNU project as another example. If someone is not happy with that, it is always possible to start their own company.
I don’t think they’re considering it a commercial endeavor, they’re just acknowledging that complex open source projects often require paid work to effectively maintain and develop them.
The GNU project works because it’s a bunch of small packages that are each maintained by approximately one person each for free on their spare time.
LibreOffice is a complex office suite that essentially competes with a multi-billion dollar industry of complex office applications and services.
It’s also an open source project that has pretty much always depended on corporate sponsorship and a paid variant rather than having some other form financial backing (e.g., it never went the Wikipedia route of being completely free for everyone and only surviving on donations).
If a text editor is not smaller than an office suite that handles Excel spreadsheets, PDFs, Word, PowerPoint, and Access databases, I made the right choice never using it.
Not really. If there are other maintainers who have ownership of the project they can just unarchive it themselves. If not, then he'd have to appoint a successor first, which would mean doing more work for free. So the best solution is just to archive it and let the community fork it if they're interested in continuing development.
Also your phrasing of "would be fine" implies that there are things that are not "fine" to do when doing work for free for the public benefit, which is exactly the sort of entitled attitude that makes many (myself included) uninterested in open sourcing their own projects.
I think it would make a big difference even if the aggregation of queries was separated out into a separate repo from the other code, so that you can pin the code for installing and updating parsers while still being able to get updates to the queries, and it isn't as much of a problem if the installation code make major breaking changes (like requiring you to use 0.12)
You just lost your job, through no fault of your own, or maybe because you did the right thing and blew the whistle on illegal and/or unethical behavior. You don't know how long it will be before you find a new job or how you are going to pay the bills until then. Your employer offers you some money to tie you over, and maybe some resources to tide you over until you get a new job, but you have to agree to a hundred pages of legalese. And you only have a few hours to decide, not enough time to have a lawyer look at it, even if you could afford to pay a lawyer. You are highly stressed, so even if you take the time to read it, you probably don't take it all in, and feel pressured to agree. And your employer also makes vague threats that about what will happen if you don't sign it, like having a hard time finding a new job or maybe having legal action taken against you.
Does that seem like a free informed decision?
Looking at it another way, anti-disparagement agreements are basically bribes to keep quiet even if disclosure would benefit the general population.
The neat thing about LLMs is they are very general models that can be used for lots of different things. The downside is they often make incorrect predictions, and what's worse, it isn't even very predictable to know when they make incorrect predictions.
reply