[9fans] plan9 finished

[9fans] plan9 finished

[matt]

this has been bugging my mind

When I said I tell people plan9 is finished I mean "complete" not "dead".

m

Re: [9fans] plan9 finished

[David Tolpin]

> When I said I tell people plan9 is finished I mean "complete" not "dead".

Do you think there is a difference between complete and dead?

Hopefully though, plan9 is neither dead nor complete, judging
by the amount of incomplete projects and intensity of development.

Dvd

Re: [9fans] plan9 finished

[matt]

>> When I said I tell people plan9 is finished I mean "complete" not "dead".

>Do you think there is a difference between complete and dead?

I prefer to think of it as stable and predictable.
The kernel is finished modulo drivers. If fairly sure there won't suddenly be a new scheduler or VM or 9p2005 or a bunch of new syscalls to surprise me.

I think people want to be sure in their time investment. If they see that someone is *still* fiddling with the kernel then if they put 6 months into learning the OS http://plan9.bell-labs.com/plan9dist/ will still be there at the end of it.

Personally I'd prefer to think that if I download it today I won't have to be downloading updates every week for the rest of my days that break all my efforts to date.

Personally I would still use plan9 if everyone else stopped, I'm sure plenty of us here feel the same.

m

Re: [9fans] plan9 finished

[David Tolpin]

> >> When I said I tell people plan9 is finished I mean "complete" not "dead".
>
> >Do you think there is a difference between complete and dead?
>
> I prefer to think of it as stable and predictable.
> The kernel is finished modulo drivers. If fairly sure there won't suddenly be
> a new scheduler or VM or 9p2005 or a bunch of new syscalls to surprise me.
>
> I think people want to be sure in their time investment. If they see that
> someone is *still* fiddling with the kernel then if they put 6 months into
> learning the OS http://plan9.bell-labs.com/plan9dist/ will still be there at
> the end of it.
>
> Personally I'd prefer to think that if I download it today I won't have to be
> downloading updates every week for the rest of my days that break all my
> efforts to date.
>
> Personally I would still use plan9 if everyone else stopped, I'm sure plenty
> of us here feel the same.

So, what is the difference between a complete and a dead?

Re: [9fans] plan9 finished

[David Presotto]

[-- Attachment #1: Type: text/plain, Size: 237 bytes --]

> The kernel is finished modulo drivers. If fairly sure there won't suddenly be
> a new scheduler or VM or 9p2005 or a bunch of new syscalls to surprise me.

Except we've had two major rewrites of the scheduler in the last 6 months...

[-- Attachment #2: Type: message/rfc822, Size: 3054 bytes --]

From: David Tolpin <dvd@davidashen.net>
To: 9fans@cse.psu.edu
Subject: Re: [9fans] plan9 finished
Date: Sat, 28 Feb 2004 17:00:18 +0400 (AMT)
Message-ID: <200402281300.i1SD0Ipm044920@adat.davidashen.net>

> >> When I said I tell people plan9 is finished I mean "complete" not "dead".
>
> >Do you think there is a difference between complete and dead?
>
> I prefer to think of it as stable and predictable.
> The kernel is finished modulo drivers. If fairly sure there won't suddenly be
> a new scheduler or VM or 9p2005 or a bunch of new syscalls to surprise me.
>
> I think people want to be sure in their time investment. If they see that
> someone is *still* fiddling with the kernel then if they put 6 months into
> learning the OS http://plan9.bell-labs.com/plan9dist/ will still be there at
> the end of it.
>
> Personally I'd prefer to think that if I download it today I won't have to be
> downloading updates every week for the rest of my days that break all my
> efforts to date.
>
> Personally I would still use plan9 if everyone else stopped, I'm sure plenty
> of us here feel the same.

So, what is the difference between a complete and a dead?

Re: [9fans] plan9 finished

[matt]

> So, what is the difference between a complete and a dead?

That is in the mind of the beholder.

I suppose dead, in the case of an OS, would mean no users whereas complete would mean active use but no further development.

m

Re: [9fans] plan9 finished

[David Tolpin]

> I suppose dead, in the case of an OS, would mean no users whereas complete
> would mean active use but no further development.

god save any software project from being complete.

Re: [9fans] plan9 finished

[matt]

> Except we've had two major rewrites of the scheduler in the last 6 months...

lol, oh well. I'm happy with plan9-unstable too :)

Re: [9fans] plan9 finished

[Brantley Coile]

Plan 9 is finished because I can use it to do stuff other than develop Plan 9.
Plan 9 isn't dead because it gets adapted to new hardare and user requirements.
I written lots of code that is unfinished and dead.

 Brantley

Re: [9fans] plan9 finished

[David Tolpin]

> Plan 9 is finished because I can use it to do stuff other than develop Plan 9.
> Plan 9 isn't dead because it gets adapted to new hardare and user requirements.
> I written lots of code that is unfinished and dead.

The original poster used the word 'complete' as opposite to 'dead'.

What you name 'finished' is 'mature'.

David

Re: [9fans] plan9 finished

[andrey mirtchovski]

> Except we've had two major rewrites of the scheduler in the last 6 months...

Is there a writeup on the changes anywhere?  I know it was mentioned
on 9fans but thought somehow that there'll be a document detailing why
the changes were made, what they were and what improvements they
resulted into, once they were complete.

andrey

Re: [9fans] plan9 finished

[David Presotto]

[-- Attachment #1: Type: text/plain, Size: 728 bytes --]

Of course not...

The normal scheduler was rewritten by rsc.  There are two
changes.  The first is a sort of ganging or processes.  If
a process wakes up another one, and the first process blocks
without using its quanta, the second gets to run.  This is to
help out processes piping to each other.

The other change was to make our scheduler to look more like
a BSD style fair share scheduler.  It was done to be nicer
to cpu hogs.

Also, sape rewrote the real time edf support to make it just
run as the top two scheduler priorities rather than a totally
different mechanism.  If there are edf processes waiting to run,
rsc's gang stuff doesn't happen, i.e., the real time processes
have priority over ganging.

[-- Attachment #2: Type: message/rfc822, Size: 2075 bytes --]

From: andrey mirtchovski <mirtchov@cpsc.ucalgary.ca>
To: 9fans@cse.psu.edu
Subject: Re: [9fans] plan9 finished
Date: Sat, 28 Feb 2004 08:11:46 -0700
Message-ID: <2e1022d55d7fb8547701ea3c5212d28e@plan9.ucalgary.ca>


> Except we've had two major rewrites of the scheduler in the last 6 months...

Is there a writeup on the changes anywhere?  I know it was mentioned
on 9fans but thought somehow that there'll be a document detailing why
the changes were made, what they were and what improvements they
resulted into, once they were complete.

andrey