Re: [9fans] p9f mention of 9front

[unobe@cpan.org]

Quoth hiro <23hiro@gmail.com>:
> > Foundation and 9front can work to figure out how to make it so there's
> > one distribution.
> 
> For 9front this doesn't matter so much. We do not have a problem with
> there being multiple (sometimes experimental) distributions. For
> non-experimental things like 9legacy and miller's rpi releases, we're
> already benefiting from each other and sharing code without any
> issues.

So perhaps I should have said 'one code base under version control',
not 'distributions'.  That is, have a master branch and different
branches for experimental things and for formal releases.  I think
getting 9front onto git is a good push forward in that direction,
since the Plan 9 is released by P9F under the MIT license from git,
too.

> > There appear to be entrenched philosophical differences, but both
> > "sides" agree that 9front is a fork and 9legacy isn't (one point of
> > view is described @ http://9legacy.org/intro.html and the other is @
> > http://fqa.9front.org/fqa0.html#0.1 ).
> 
> I disagree here.
> 
> When 9front was created it was technically a fork, because there was
> still a diverging bell-labs plan 9 distribution.
> 
> We may seem to have new non-accreditated "management", but we are all
> fascinated what Plan 9 brought to this world and we would all like to
> build on the general wisdom of the Plan 9 philosophy.
> 
> Now that bell-labs is gone, there have been no more mainline plan9 releases.
> 
> Instead we have a few remaining relevant distributions like 9legacy or
> miller's rpi releases, all of which technically are about as much a
> fork as 9front is.
> 
> Erik's 9atom has been unmaintained for a longer time so we didn't have
> any problems following those patches.
> 
> Harvey and related distributions are highly experimental, they diverge
> much further from Plan 9 basic architecture without any will of
> keeping backwards compatibility.
> They are more revolutionary than some of us can stomach.
> 
> A lot of people from the old crew at bell-labs completely abandoned
> mainline bell-labs plan9 even before 9front has been started because
> they seem content with just having a p9p layer on top of their
> macbooks or other unix machines.
> 
> Apart from the occasional trolling that keeps on coming up on this
> list, I don't see what deep trenches people are imagining would impair
> the Plan 9 Community from working together on a well maintained Plan 9
> distribution with simple procedures for sharing code in all
> directions. Whose wrong foot did you suffer from when 9front got first
> made?

For one, not having fossil.  The removal of fossil might be what some
users cannot accept?  The setup and wiki preserved at 9p.io talk about
how to setup fossil+venti and yet that cannot be done on 9front--it
was removed completely.  I think I recall you saying it was bug-ridden
and unmaintainable, but could be misremembering who said what.  From
what I can gather, there was a serious bug that was fixed in fossil
after 9front forked, and yet there is no intention of including it
back into 9front.

Additionally, I didn't realize initially that there was only one
kernel for both cpu and terminal in 9front, as opposed to two in Plan
9.  (I like that there's only one, personally.) That makes a
difference because of the documentation in each system on how to setup
a server.

Those are two things that I can think of off the top of my head.

Philosophically, I think the 9front maintainers/developers are much
more willing to chuck older code and to replace with new code.  I have
personally benefitted from this approach so I'm not against it.
9legacy is much more conservative.


------------------------------------------
9fans: 9fans
Permalink: https://9fans.topicbox.com/groups/9fans/T523d6e906a17a7cc-Ma33a7ff5d2238a0c5a82a824
Delivery options: https://9fans.topicbox.com/groups/9fans/subscription