Re: Linux headed for disaster?
Iusty Pop Daniel (iusty@netcompsj.ro)
Mon, 6 Dec 1999 11:09:29 +0200 (EET)
What I don't understand is: if linux is so unstable (quote from an
early email from Kendall on this thread), if linux is doomed because of
lack of binary modules, why does Kendall need linux so bad? He says that
OS/2, NT, and others are much better because of x and y and so on, so why
does he still uses linux? I don't understand, I cannot complain about my
linux being unstable, compared to win (be it 9x or NT), how does he?
Btw, I use redhat, but first thing after I install a new linux
somewhere, I replace their kernel with the latest one, compile it myself,
and never, EVER, got any problems because of this. I have servers running
like this for 2+ years now, with upgrades from libc to glibc2.0 to
glibc2.1 made by hand, and no, sir, no problems with compiler version or
something else. Sometimes I get the src.rpm and compile it myself and no
problems there either.
And if you want to know, I use on a server a Telemann SkyMedia 200
sattelite receiver card, whose drivers comes as a x.c and y.o, compile
x.c, link it with y.o and you have the driver ok. Hardware details are in
y.o, but kernel interface is in x.c, recompilable under (most or all)
2.2.x versions. But I had a problem with the card, and while I could trace
it in x.c, when it called a function in y.o, I was lost. I even tryed to
objdump -d y.o >y.s, but I'm not so good at assembler to understand such
assembler code. I was glad that there was x.c, but sorry that there was
y.o. Anyway, that solution works, only having to recompile x.c at a kernel
change, isn't that some kind of binary compatibility? With no extra
something, only standard kernel tree? And, personally, I would NOT use a
module from 2.0.x or 1.2.y, even if I could.
Iustin Pop
-
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/