Re: Future of m68k - Etch and beyond
* Ingo Juergensmann:
> I still have an i486sl33 machine at home. It feels faster than my 060
> Amigas, although the 060s should be superior to the 486. So, maybe there's
> an arch specific slow down (BE/LE)?
Thats absolutely true, with the introduction of m68k Debian (2.0?) I
made some test and basically it can be sumed up as following:
GCC produces really mediocre code on m68k
m68k-code is bigger while m68k systems have less memory
Debian uses rather generous options on its libs etcpp - which eats up
even more memory and cpu-cycles. Eg when I dropped locale-support and
some other "nice-to-have-but-useless" from glibc/xlib/etcpp (remember,
that was libc5/xfree3, todays memory-bogs glibc2.x/xorg7 are even more
prone to this) suddenly even m68k felt very fast.
All in all I wouldn't be surprised if a lean m68k system GUI would feel
three times faster and would only need half of nowadays memory. Because
m68k has nearly nothing to do with nowaydays requirements of Debian-Targets.
>> [lots of stuff deleted, because I'm in a hurry ]
>> I would prefer to see Debian continue to support the m68k architecture.
>> But if it simply isn't going to happen anymore, I really think that a
>> more focused effort,
We shouldn't focus too much on the name "Debian" but more on
"Debian-based-Core". In my oppinion is shouldn't be Debian dropping m68k
but m68k dropping Debian. Take the core, develope a very-debian-close
small distribution and voila, we are done.
Christian Brandt
Reply to: