[Sussex] Debian news...

Steve Dobson steve at dobson.org
Wed May 4 10:58:06 UTC 2005


Geoff

On Wed, 2005-05-04 at 10:15 +0100, Geoffrey J. Teale wrote:
> Steve Dobson <steve at dobson.org> writes:
> > Geoff: Do you know why older RISC CSUs are being dropped?  Is it because
> > they can't find working hardware on which to test, or is it because
> > those CPUs are not in common use?  The former is understandable, but if
> > it is the latter, then that is a really shame.  The FSF will be talking
> > away my freedom to use a paricular CPU.
> 
> Well, pretty much it's a combination of both of those factors and that
> maintaining those (very obscure) architectures has become a stumbling
> block in the road to a better compiler.  The phrase RISC CPU's above
> was misguided.  I've reread the release notes and to be clear GCC4
> drops support for the following chips:
> 
>     * Intel i860
>     * Ubicom IP2022
>     * National Semiconductor NS32K
>     * Texas Instruments TMS320C[34]x
>     * SPARC family
>           o SPARClite-based systems (sparclite-*-coff, sparclite-*-elf, sparc86x-*-elf)
>           o OpenBSD 32-bit (sparc-*-openbsd*)
> 
> Which seems fairly unlikely to hit most peoples freedoms.  Of course
> if you need support for those things you can always continue to use
> (and even maintain) GCC 3 - which is what Free Software is all about.

Absolutly, although I would say it is _one_ of the things that Free
Software is all about.

But playing Devil's Advocate here one of the "selling" points of Linux
(and F/OSS in general) is its support for older, legacy hardware.  For
commerical, propriety software you get support as long as you pay the
price.  Microsoft support old releases of Windows much longer than
Debian does, giving users much longer to upgrade.

When sarge becomes stable then security fixes for woody (current stable)
will have a very limited life.  If you're using woody at the moment
start planning your upgrade.

The point I am trying to make here is that I don't think the ecconomics
of Free Software is fully understood yet, and as a result we are still
following (broadly) the same procedures as we did with propriety
software, but for different reasons.  So are we any better off?  Well I
think we are, but it is still important that we ask the question, and
keep asking it.

> > Now I suppose that some might say that I am free to add support myself,
> > which it true.  But one of the strongest arguments for using high level
> > lanugages is that you get platform independance - they serperate you
> > from the underlying hardware.  I can develop software for any processor
> > for which I have a C/C++/... compiler.  I'm not an expert in either
> > compiler tecknology or any CPU, so to add support to GCC4 for a dropped
> > CPU I would have to re-train myself.
> 
> I see you preempted me :-)

I thought it would save some time. :-)

> > The FSF, by supportting CPU-x in GCC<=3, are taking away a freedom that
> > they gave me in the first place.  I'm not saying they don't have the
> > right to do that, and I accept that they may have to do this for
> > financal reasons.  I just think it goes against what the FSF stands for
> > it they are doing this for commerical/market type reasons. 
> 
> No it's not about commerce or finance, it's about delivering a better
> compiler.  They're also not taking away GCC3 so in the unlikely case
> you need it then you can use it.

This is one really big advantage of Free Software, the fact that the old
technology (GCC3 [and 2 and 1 and ...) are still available.  If you look
at it this way I haven't lost any of my old freedoms, I'm just not going
to get any new ones on old CPUs.

Steve





More information about the Sussex mailing list