From mboxrd@z Thu Jan  1 00:00:00 1970
From: Corey <corey@bitworthy.net>
To: 9fans@9fans.net
Date: Sat, 17 Apr 2010 02:39:28 -0700
User-Agent: KMail/1.13.2 (Linux/2.6.34-rc4; KDE/4.4.2; i686; ; )
References: <2732228b893d9ead6b8e1cb3d2e698b1@proxima.alt.za>
In-Reply-To: <2732228b893d9ead6b8e1cb3d2e698b1@proxima.alt.za>
MIME-Version: 1.0
Content-Type: Text/Plain;
  charset="us-ascii"
Content-Transfer-Encoding: 7bit
Message-Id: <201004170239.28944.corey@bitworthy.net>
Subject: Re: [9fans] Mars Needs Women (was Re:  TeX: hurrah!)
Topicbox-Message-UUID: 057d2ba0-ead6-11e9-9d60-3106f5b1d025


I appreciate your time and consideration in your responses, thanks!

You made several points and asked several questions this email,
however it's difficult for me to answer them because they appear
to be put forth under the idea that "Plan X's" purpose is to natively
host common popular consumer-level, end-user applications
of various sorts under Plan 9, and/or to port gnu to Plan 9.

There also seemed to be a lingering impression that I'm
suggesting that Plan 9 proper - the official distro - should be
subject to the changes proposed by "Plan X".

However I think it's crucial that the current official Plan 9 distro
continue as it always has. And I don't personally see enough value
in the notion of gnu gcc and autotools being ported, or firefox and gtk,
etc..

I do see  value in porting LLVM/Clang, which would help enable,
for instance, a forked  and customized _subset_ of the EFL core
libraries (not the E wm), ported and running like a native Plan 9
citizen via /dev/draw instead of X.

I'm imagining an alternative Plan 9 distro that jettisons just a couple
select characteristics of the system which drastically increase the net
sum total alien'ness that tends to obfuscate and/or divert attention
away from (what I believe to be) the more important aspects of the
Plan 9 experience, such as the ones I listed previously:

* 9P
* mutable namespaces
* union directories
* ubiquitous fileservers
* transparent distributed services

Slightly more POSIX - but not total POSIX compliance - in
addition to a non-gnu compiler that supports modern standard
C dialects and other C-based languages would be an enabler
for a greater number of people hoping to apply Plan 9 concepts
under a broader and more general variety of purposes.

But even all that begins to miss the original attempted point of my
first post: the idea that perhaps it could be beneficial if there
were some means for interested Plan 9 fans to rationally discuss and
speculate on different potential expressions of Plan 9 based operating
systems.

Attempting to do so here on 9fans continues to be a traditional source
of agitation and flames, tempered with a healthy dose of shut up and
code. (that's not an accusation or scornful judgment, just a statement
of a "thing").

I thought that perhaps talking in terms of what "a 'Plan X' _might_ look
like" would be less divisive/threatening than talking in terms of what
"Plan 9 ought or ought not become".


Cheers


On Saturday 17 April 2010 00:41:19 lucio@proxima.alt.za wrote:
> > ... are simply - by far - much more important and practical to a greater
> > number of people than these other prominent Plan 9  idioms:
> >
> > * radical frugal simplicity throughout the entire system
>
> This would remove itself as soon as the developer base increases
> beyond an indeterminate critical mass.  That's precisely how Linux
> grew beyond Minix.  But there is Linux already out there, so no
> clarion call to developers to move to a less popular platform.  Plan 9
> and NetBSD have many philosophical issues in common and both suffer
> (benefit?) from a shrinking user base because populism (fashion)
> rules.  Polluting Plan 9 with fashionable toys isn't going to save the
> world, isn't even going to be useful to the existing Plan 9 community,
> so why do you believe it should happen, rather than allow Plan 9 as it
> exists, both as a philosophy and as the implementation of this
> philosophy, to demonstrate that a simpler lifestyle is also
> sufficient?
>
> What do you see in a "liberated" Plan 9 that would make it superior to
> the existing tools out there?  Or, to ask the same question in a
> different form, why do you pick on Plan 9 to become your target
> platform through unwelcome (*) transformations instead of transforming
> that which is already much closer to your objectives?
>
> (*)	"unwelcome" both because some of us believe it to be ethically
> 	undesirable and because the more pragmatic ones amongst us have
> 	not found sufficient motive to focus on them.  Take fgb, for
> 	example, who found cause to port curses to Plan 9, opening the
> 	door to many new developments: few have done much with this, what
> 	changes would you effect that would increase these contributions
> 	significantly?
>
> > * a stance against POSIX and other standards
>
> The stance is against polluting Plan 9 with inconsistent,
> committee-defined functionalities that often contradict even common
> sense.  Posix is yet another cesspool where nothing is ever removed,
> no matter how foul.
>
> > * a stance against alternate programming language paradigms
>
> Not at all, only against extending C in a direction that has been
> shown to be counter-productive.  Alef was dropped out of necessity,
> Python and Perl are available, Go has been considered, Tcl was ported
> moderately easily, it is only the G++ model of C++ that has been
> proved intractable.  Sadly, that is what everyone is clamouring for,
> so it looks like a much bigger issue than is truly the case.  The
> problem here is again not of Plan 9's making, it is that the mass of
> developers have no understanding of portability and therefore paint
> themselves into the Linux corner.  Again, how do you propose to alter
> Plan 9 to address this form of antisocial behaviour?
>
> > * a strong bias towards a particular form of user interaction with the
> > system (i.e. acme, rio, etc)
>
> There are already two camps in Plan 9, one that uses acme, the other
> uses sam and many experiences cannot be shared because of that.  Are
> you sure you'd improve on this by releasing hundreds of customised
> window managers for people to share even less?  How would adding emacs
> as an editor improve matters, as an example?
>
> Now, imagine that in your Plan X context somebody actually ported
> Firefox: what kind of gymnastics would it take to feed the port
> upstream and make sure that the next release does not destroy all the
> efforts?  And what are the chances that the various extensions to
> Firefox would also be ported and maintained?  Where are you going to
> find the good will and resources to maintain just one of this class of
> projects, nevermind the tens of thousands out there (of which GCC/G++
> is one, by the way, why is it so hard to port it to Plan 9, if it is
> such a portable piece of engineering?).  And, most crucially, why
> would anyone offer to do that when it's already available?
>
> Linux filled a gap by being free at the time when there was a great
> demand for inexpensive and unencumbered software to match the
> ridiculously low price of computer hardware.  No analogous demand
> exists today that would be satisfied by the Plan X you envisage, or,
> more humbly, perhaps you can show me what such a demand is.
>
> But if the demand is, as is my case, for a simpler, easier to
> maintain, easier to understand computing platform, then Plan 9 and not
> your Plan X, is the answer.
>
> ++L