This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: glibc stacking up to Windows
Ulrich Drepper wrote:
[I wouldn't normally reply to nonsense like this but the record should
be straight once.]
Keep calling it nonsense, M$ will keep winning the war. Come on guys, someone (us) actually trying to provide code in a cross-platform way, thus we need equiv API's from both OS's, and it has to
result in war from both sides. Both sides are equal idiots.
Just use posix_spawn() to run cp. It makes absolutely no sense to
replicate the functionality of cp. This is no simple program, it's a
high complex and optimized one. Duplicating its functionality would be
an invite for errors and missed opportunities.
Why leave it only as a program to run... MIGRATE the code into a lib, transition cp simply to being the program which accepts and validates user switches, and interfaces with the lib. Now you have
equiv code being able to be called by other programs via the API and cp at the command line. You don't end up with duplicate functionality once the migration is done.
Why do we need to take the performance hit to start another PID? Remember CGI that got the bum wrap about performance for starting PID's all of the time? I would want our program to get an equiv wrap
why exactly?
Before you take to your keyboard the next time and complain about
something you obviously have no idea about read up first. Read about
the POSIX project and think about what all this means.
Oh I know about it, and this BS is what I expected. It stinks, it absolutely stinks.
"It's always been this way. We ain't changing it. That's not the goal of this project." bla bla bla.........
May you all be pleased with loosing the war!
--
Michael Lueck
Lueck Data Systems
http://www.lueckdatasystems.com/