Re: Proposal: remove OSF/1 support

[Greg Troxel <gdt%lexort.com@localhost>]

"J. Lewis Muir" <jlmuir%imca-cat.org@localhost> writes:

> You might have guessed that I would push back on this a little. :-)  I

Actually I had blocked the previous conversations out of my mind and
didn't guess.

> haven't read the Reddit rant, so I'm not talking about that at all, but
> I do think this idea of "here's some stuff I'm hacking on, but it's
> broken all over the place and not really useful, but you're welcome to
> help" is not always what users have in mind when they're using open
> source software.

They might not, but that's the way it actually is.  There are a lot of
people that think open source is just like if they had bought a support
contract, except they didn't.  At least that's how some people behave.

> But there are other users, and it might even be some of the same users
> who wanted to hack on a project but in a different context, who are
> wishing to use open source software in a production environment (at work
> or home), and they need it to be secure, reliable, and to not break
> backward compatibility.

So the only point that I'll agree with on there is that it's good to
describe if things work or if they don't.

The idea that someone is running OSF/1 on a computer and sees that
pkgsrc says it has some OSF/1 support and then formulates a plan to
isntall lots of things and use it in production is beyond insane.
Besides that anyone engaging in retrocomputing should know 1) that their
entire OS is dicey and 2) that OSF/1 is no receiving security updates
and thus should not be seriously used, anyone planning to use software
for anything serious should do some kind of evaluation and acceptance
testing.  It's not our job to save people from themselves.

> Now right there, you may say that that's not part of the deal, and users
> need to either contribute themselves or pay someone to contribute.  And
> I hear you, but I think there's at least a third model here: users who
> maintain their own open source projects and want to use others.  So,
> it's kind of like economic specialization: instead of everyone growing
> their own food and making their own clothes, etc., they specialize and
> then trade.  IMO, the same thing can happen in the open source software
> world where a user might maintain their own open source project but
> then wish to not have to become an expert in someone else's open source
> project; they wish to just use it.

Sure, that happens a lot and I do that all the time; I am a maintainer
of a few things, a contributor to a bunch, and a non-contributing user
to many.

It's a great thing in the open source world that we can all use things
built by others, and because copying has essentially zero cost, we are
all better off.

But when something doesn't work, I'll file a bug that it doesn't work,
which is a step that is helpful.  I do not act entitled and tell people
they have some moral obligation to help me in particular.

> So, I think I said it in our last discussion a few months back related
> to a security issue with OpenSSL, and I know you're looking at it from
> more of a legal point of view, and I'm not (for better or for worse),

It would be good if you understood the legal issue.  People who do are
offended when others wrongly assert that they have a duty.

> but I still feel like there's some basic level of responsibility in
> maintaining an open source software project that the maintainer should
> take on, and I think that includes ensuring the software is secure,
> reliable, and not breaking backward compatibility (within the same
> major version number); and of course if it's pre-1.0.0, then backward
> compatibility is not promised.

I continue to find your assertion of responsibility to be offensive.  I
think it's good for people not to misrepresent things, but an assertion
of duty beyond non-misrepresentation is a real problem.

> Now, as for this OSF/1 situation, I'm OK with nia's change to
> (presumably)
>
>   https://www.pkgsrc.org/
>
> to add an "other supported platforms" section and wording to make it
> more clear that it might not work.  Personally, I'd go even further and
> break the platforms down into tiers just like NetBSD does at
>
>   https://wiki.netbsd.org/ports/
>
> And with zero knowledge of OSF/1 and pkgsrc on it, but just based on
> what was said in this thread, it seems to me that OSF/1 would go into
> "Tier III: Life Support -- severely incapacitated or broken." :-)
> Just my opinion, of course.  And I'd be willing to contribute such a
> breakdown to the pkgsrc website if you like the idea.

If you had read bootstrap/README you would see that there is already a
bunch of categories and text.  But I forgot that it was there, and I
probably wrote it.  I added a pointer to that from the web page.

I have added a * to OSF/1 to indicate that it is currently believed
broken.

Things only happen when Someone(tm) does them.  So if you want to look
at that list and try to run pkgsrc on various platforms and send a
report to pkgsrc-users yea or nay, I'm happy to edit in a date-stamped
works/doesn't-work to the per-platform README.

If you don't want to do that, that's of course your call and I won't
criticize you for that choice.  But please stop demanding that other
people do things like this.