Re: Linux 2.4 before 2001?

From: Magnus Danielson (cfmd@swipnet.se)
Date: Thu Jan 06 2000 - 15:11:20 EST


From: Rik van Riel <riel@nl.linux.org>
Subject: Re: Linux 2.4 before 2001?
Date: Thu, 6 Jan 2000 15:38:39 +0100

> On Thu, 6 Jan 2000, Magnus Danielson wrote:
>
> > Smaller steps. Maybe one should open up a new development path as
> > one say "freeze" on the current so that one can redirect requests
> > for the next development tree and have those efforts start early
> > in the branch.
>
> Smaller steps is a good idea, but you cannot really split off
> a new development branch before you've stabilized the previous
> one into a stable release. The reason for this is that you need
> to have a stable development basis in order for your "small steps"
> idea to work at all...

Yes, there are troubles involved with doing that. But it is also a way of
redirecting people away from what you are trying to stabilize and basically
say "OK, here you can add code in an early stage, but be prepared to fix it
as we others come along". Thus, it is more a fashion of getting somewhat wider
spreading at an early stage for those so interested than anything else. It is
not needed to acheive a higher release frequency, it is just one fashion of
going about and help doing it.

The smaller step approach is basically say that we are trying to acheive less
new fundamental changes per release. Some are new and some are redesigns.
The methods of getting there migth be diffrent, and opening an new development
branch is just one possibility (which may also proove totally unecessary).

I beleive that up to some point, making too large steps in advancing the kernel
will make the update-and-stabilize part hard and long. Keeping the steps
smaller can aid in quicker update-and-stabilize phases. This is true only up to
some point and I am not saying i can judge exactly where this point is.
In the end we are talking about available human resources and how those are
being managed. This is a difficult task.

Either way, there's the constant figth between balancing quality and release
time/frequency.

I see no reason to lower quality, rather the oposite, but we should try to
do something about release times maybe.

Cheers,
Magnus

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Fri Jan 07 2000 - 21:00:07 EST