Text 5021, 68 rader
Skriven 2006-01-17 16:01:44 av Kenneth Parrish (8:8/2)
   Kommentar till text 5015 av Maurice Kinal (1:261/38.9)
Ärende: Re: gcc 4.1
===================
-=> In article 17 Jan 06  12:46:56, Maurice Kinal wrote to Kenneth Parrish <=-

g'day :)

[ developmental mini-systems development ]
 MK> No rest for the wicked.  I am just looking over the last successful
 MK> build using gcc-4.0.2 and editing the script(s) used to build it.
 MK> Overall I am happy with how that worked out despite problems with
 MK> certain popular packages, such as kde, which doesn't appear to be
 MK> bothering me personally.  What I should do is bring what I have in
 MK> sync with slackware, primarily pkgtools.
gcc 4.0.2 is the latest i've seen, and i do have a copy here from a friend's
having lent me the use of their internet connection, and made a mini-cdr with
it for me. from playing round agin yesterday with the some of the special
processor configurations in gcc-4.0.2/gcc/config/i386/i386.c and compiling just
the 'c' component, and installing having specified
--prefix=/usr/local/gcc4.0.2 so everything goes to
its own directory tree thus allowing easy removal and concurrent use of my
slackware packaged versions, (this sentence is getting too long), hmm, it works
fine in many areas, but shows regressions in others, according to a couple of
benchmarks recompiled with
-march=i586 and -O2, i think, from this and other
tests, it's fair to conclude gcc c compiler through v4.0.2 has never produced
optimal code *always* for the cyrix mii. but, i don't know what tradeoffs there
may be, so perhaps it's the best compromise. and i don't have the information
to best change the machine description and i386.c files to better model this
processor's variations from the pentium and pentium pro (it seems to be
somewhere in between, with elements of the pentium II and i486).

 KP> [ Re: f9x ]
 KP> Q. the two oldest high level languages still used much? A. fortran
 KP> and lisp
 KP> [jargon file]

 MK> I know why fortran but not lisp.  I have run across a few people who
 MK> still use lisp but they seem rather rare.  I suppose the same could be
 MK> said about fortran.
there's a funny acronym for lisp, language with irritating superfluous
parentheses. well, gd&r maybe.

 KP> Re: slack jsdk demos, portable perl there are several things like
 KP> that, got a start elsewhere

 MK> I don't know but have run across microperl in a few embedded
 MK> and crosscompiling projects which is what originally peaked
 MK> my curiousity about it.  Works great for the basic builtin
 MK> functionality of perl so if one is happy with that then it works
 MK> fantastic.  I noticed that when compiling X the source there seemed
 MK> to have no problem with microperl when using it to format.  X also
 MK> compiled without any perl, micro or otherwise.
there may be a couple of font conversion steps that dont complete without perl,
i dunnow.

 KP> Re: windows exploit 'back door' ?
 KP> who could believe it!?

 MK> We'll see.  So far I've only heard what was said on the news about
 MK> it, which wasn't much.  If I had a business that relied on MS
 MK> products I think I'd be talking to my lawyer about it.  Definetly
 MK> something weird going on there.
it's something how many ways there are to overload a buffer, or something, and
get into a system. who knew?

... You are worth your weight in gold, or even in the new silver coinage !

--- BBBS/NT v4.01 Flag-5
 # Origin: FamilyNet Sponsored by http://www.christian-wellness.net (8:8/2)
 * Origin: BBS Networks @ www.bbsnets.com 808-839-6036 (1:10/345)