Shortening the Development Cycle...

Robert Dinse (nanook@eskimo.com)
Thu, 9 Sep 1999 02:34:19 -0700 (PDT)


>
> On Wed, 8 Sep 1999, Carling <m@idiom.com> wrote:
>
> On Wed, 8 Sep 1999, Bill Rugolsky Jr. wrote:
>
> > As the development cycle shortens this will become more feasible.
>
> The Linux development cycle has not been shortening. I would love to see
> it shorten, but I'm not holding my breath. I've given up on hoping that
> 2.4 will be out faster than 1.0, 1.2, or 2.0. I'm still hoping it makes
> it out faster than 2.2 did.
>
> > the reality is PC hardware changes constantly: ACPI, USB, ..., the list
> > is very long. Are you really suggesting that we only support new
> > processor variants or chipsets (like Athlon or the Intel 810/820) in
> > the development kernels until the next stable release?
>
> Yes. Such a change would shorten the development cycle because it would
> change the pressure on Alan and Linus from delaying feature freezes to
> enforcing feature freezes. I think many features would become available
> in the new stable kernel series faster than they would have been back
> ported to the old series. It would also save much duplication of effort,
> which I think could be better put to use on new development.
>
> M Carling

It would limit participation in development to those that can afford the
latest and greatest hardware, significantly reducing the pool of coders, and
as a result increase the length of the development cycle.

It would result in code that would only run on the latest and greatest,
and many things would be incorporated into the code that does not work on the
majority of platforms but hasn't been tested, until the freeze, then you would
have this giant mess to clean up, that at that point may not even be possible
to clean up, lengthening the development cycle.

One of the things that initially drew me towards Linux was the relatively
large number of hardware platforms supported. Of all the things I hate about
Windows, the fact that it runs essentially only on one platform is high on the
list.

I have to admit that I am disturbed by the influence that M$ thinking
seems to be having on Linux as of late, new interfaces that look like M$'s
interface, not only in the OS itself, but in the installation programs and
interfaces shipped with the commercial distributions, and now the attitude that
yea, just like M$, we're now going to force you to throw out your hardware and
applications every year else you can't enjoy the latest wiz-bang gadgets.

Such a move I'm afraid would result in a good deal of elitism amoungst
those that can afford to participate, resulting in an OS that appeals to a
snooty few who can afford the hardware to run it, and not one that appeals to a
diverse userbase like that which presently exists.

-
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/