Re: [9fans] interesting potential targets for plan 9 and/or inferno

[bride of excession <bride.of.excession@gmail.com>]

On Wed, 7 Mar 2007 06:43:47 GMT
rminnich@gmail.com (ron minnich) wrote:

> On 3/6/07, Vester Thacker <vester.thacker@gmail.com> wrote:
> 
> > Well, I apologize for offending you. But what ideas are you
> > suggesting that we migrate toward?
> 
> no, sorry, I was kind of short. No problem :-)
> 
> I actually agree with your points in many ways. I just don't know how
> to get around the problem of showing this system to people. It's such
> a powerful system, and it drives me crazy when the first reaction is
> "I don't like that GUI". I'm talking BEFORE I've typed anything in the
> little window. These are not dumb people. But they have work to do,
> and they don't see that climbing the learning curve is worth it.
> 

For me it's not the learning curve, it's the fact that my eyes hurt
immediately from looking at all the brilliant white light. And then
there's that quirky toy-like mouse business... What's wrong with using
keyboard shortcuts anyway?


> Then I come to this list and people say "fuck 'em". I don't know,,
> seems like an unconstructive attitude to me.
> 
> There's lots of stuff missing, as I pointed out at other times, in
> other notes. A lot of things that are missing are needed to get
> continued $$$ to keep things going. As jmk pointed out some time ago,
> plan9.bell-labs is supported in part by DOE, but at some point, if we
> can't show certain things, then the money goes away. The recompete
> happens this summer.
> 
> The isses of Python and gcc are not simply academic. They're part of
> the DOE meal ticket.
>

Yes, gcc and python would go a long way toward making plan9 useful in
the real world (vs only the research world or only the embedded world).

I see a lot of plan9 people critical of the idea of "copying windows"
yet that funtionality is precisely what practically everyone
(including plan9 folk) needs -- the paradigm is already cast and user's
standards (of a sort) *are* for all intents and purposes writ in stone
simply because we *need* to edit those .doc files, read pdf's and watch
DVDs. For crying out loud, there is *still* no graphic web browser and
it's the year 2007! 

New paradigms don't come from casting the old ones aside, rather they
evolve from the old pardigms. But that can't happen if one is so 
"idealistic" as to find compromise unacceptable. 


> Users matter ... I've just gone through an interesting few years of
> fielding the highest performing cluster software anywhere, and having
> people get upset because they can't run emacs or xterm or gdb on each
> and every one of 1024 cluster nodes. The system we have starts up
> 2048-cpu MPI jobs in 3 seconds, and the competitors literally take
> minutes (one recent system). It boots 1024 or 2048 nodes in 2-4
> minutes ,and the system replacing it can take hours -- yep, hours --
> to boot. This is all without a local disk, mind you. But people want
> that ssh login ... and /etc/passwd ... and xterm ... and so on. it's a
> conservative world in computing, nowadays. Everybody wants everything
> to look like a linux desktop, even a cluster node. It's kind of sad.
> Clusters are stuck in a 1997 mentality.
> 
> So, yeah, users can be frustrating, but they are your meal ticket.
> Saying "fuck em" only works just so long -- as you may have noticed,
> most of the Plan 9 guys are at google, running large Linux clusters
> ... and I believe many of them carry macos systems around now.
> 
> ron


Because "waiting for the new paradigm" just doesn't work unless one has
a clear vision and plan for achieving it. Sad.

But anyway I don't mean to barge in and pitch a bitch, and I certainly
have no plans to abandon plan9 over any of this, I'm just a tad
frustrated.

:) 

b.of.e