Port-i386 archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: 80386 support

> As for supporting sun3, i386, etc - perhaps you have a wrong
> impression.  It was stated before that NetBSD is not a conservation
> project nor its primary goal is to run on the old computers

There was a time when NetBSD said "Of course it runs NetBSD." and meant
it.  It was an explicit goal of NetBSD to run on old computers, new
computers, esoteric computers, common computers - pretty much anything
with an MMU and at least 32 bits of width in its programmer model.

That is the NetBSD I joined.  That, or something like it, appears to be
the NetBSD the acquaintance I quoted remembers.

He wrote of the three-tier paradigm as being a huge change.  It was,
and is, as compared to historical NetBSD attitudes and policies, even
though by the time it came in it was just a formal recognition of the
attitudes that had already crept, step by tiny step, into the project.

> There are other active projects (e.g. 2.11BSD revivals) which have
> such goals.  Don't they suit your needs better?

Possibly (though I think 2.11 was for 16-bit machines).  I need to
investigate such things more; if I find one that strikes me as better,
I will switch.  I'm still here for four reasons:

(1) Inertia.  NetBSD is what I've run, almost exclusively, on my
machines for some 15-20 years.  The activation energy needed to change
that, the degree of mismatch necessary, is relatively large.  And the
search for "something better" is a tedious and difficult job,
especially for someone with as many different ports as I have.

(2) For all that is an increasingly bad match for me, NetBSD is the
best I've found so far - the best of an increasingly bad lot, you could
say.  (Gradually-getting-older NetBSD; I haven't yet had reason to push
any of my own machines to 5.x, and don't consider 6.x an option at all,
for reasons I've ranted about elsewhere often enough.)

(3) I've been working on and with NetBSD for 15-20 years.  I've formed
a certain amount of emotional attachment to it.  If a parent sees a
child doing something self-destructive, it's difficult to not care
about it, to not speak out against it.  While I suspect the emotions
are weaker here (I have no children of my own, so i can't directly
compare), I think something along those lines is operating.

(4) I can still help people here occasionally.  For all that NetBSD is
going in very wrong directions for me, there are a lot of people here
who have helped me tremendously in the past; I feel I owe it to
everyone to pass along a little of the help I've gotten.

None of these will keep me here forever, of course, neither
individually nor collectively.  If NetBSD persists in walking the path
it's been turning down, sooner or later I'm going to have to leave.
I'm already effectively maintaining my own fork, in the form of
significant private patch trees; I may simply call it a fork.  Or I may
switch to something else, if I can find something.  Or I may get out of
computers altogether.  Or who knows....

> However, if MI code is too hungry on a system with 32MB of RAM, then
> I think we ought to investigate and fix or improve that.

Easy enough to reproduce: install NetBSD/i386 (4.0.1 in my case; I
don't know what other versions might provoke it) on a 32M x86 system
and do a build.sh build.  I forget the exact hardware I used, probably
soemthing in the PII or PIII range - I doubt it matters.

> Anyway, given the described maintenance problem with acorn26 and
> sun2, you expect somebody to write and more importantly maintain BPF
> JIT for these ports?  Who will do that?

For "expect" meaning "consider likely", no, I don't expect that.

For "expect" meaning "consider reasonable or due", actually, I expect
BPF JIT to stay out of the tree until it is substantially more mature
and supports all NetBSD's ports, and I expect it to be the duty of
those arguing for its inclusion to do, or arrange to get done, any work
necessary for that to be true.  But then, by that definition, I expect
a lot of things which NetBSD has already failed totally at.

> Do you expect others (i.e. not those few who might actually be using
> them) to do it?

No; NetBSD has made it clear that those using other than first-class
ports should not expect anything except crumbs happening by chance to
fall from the table.

Yes, that's emotionally loaded language.  That's because that's the
emotional reaction I have to being told that NetBSD no longer cares
about the goals and principles it had when I joined, and for most of
the time I spent working on it.  In my (apparently worthless) opinion,
modern-industrially-relevant-BSD should have forked off as a separate
project, rather than co-opting the NetBSD name - and the goodwill it
built up - for a project that is drastically different from the project
that earned it.

And if you think emotional reactions don't matter, you're forgetting
that NetBSD users are humans...and humans are emotional creatures.

/~\ The ASCII                             Mouse
\ / Ribbon Campaign
 X  Against HTML                mouse%rodents-montreal.org@localhost
/ \ Email!           7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B

Home | Main Index | Thread Index | Old Index