hch at infradead.org
Thu Jan 3 16:16:52 PST 2002
On Fri, Jan 04, 2002 at 12:22:17AM +0000, Alan Cox wrote:
> Well if you don't want any users its not my problem. When you've finished
> writing calderux with its own libc
[Sender address changed to stop your silly assumptions.]
Hurd uses the same glibc, that's not the point.
What I mean is that we there is no reason to support a feature
forever just because glibc (ab)uses it. /proc is not set in stone.
> that runs no normal Linux apps you, that runs no normal Linux apps you
> can go and join the hurd team.
IF LSB would be designed in a sane way there would be no reason to not
to do so. But as it contains numerous random glibc internal symbols
it gets rather difficult to use a different libc.
Same is for specifying all those GNU options in the tools, that's very,
very bad style.
> Real world customers (the sort who pay sensible amounts) expect and demand
> forward compatibility and the ability to roll back cleanly. Even if Linus
> were to take it out of his kernel it would be relevant to the LSB as all
> the vendors will have to put it back compatibly.
Glibc already contains so much version checking code that it would be no
problem to use /proc/cpuinfo if the kernel version is smaller than
foo.bar.baz and a sane interface if later.
Of course it doesn't work. We've performed a software upgrade.
More information about the lsb-discuss