Re: fork() and progress bar...



On Thu, 06 Sep 2001 18:41:30 CDT, Michael Lee Yohe <myohe redhat com>  said:

> The same reason "our" hello world programs look different.  To each his own.

You are in a twisty little maze of API's, all different.  (with apologies
to Crowther and Woods).

There's a reason to avoid using a slightly-different API - support costs.
If gthreads had a different thread model than Posix pthreads, it would be
one thing.  But requiring the support crew to learn pthreads *and* gthreads
when there's no real advantage to it is just wasteful.

Heck, I know *I* would have spent a heck of a lot of time trying to figure
out what subtle differences, timing holes, etc gthreads had, and what the
gotchas are, and why the original author didn't just use pthreads instead.

Ever had to debug a problem because somebody used the memmove() calling
sequence for bcopy()?  I have.  It's not fun, especially when it's buried
in a signal handler that only trips under some conditions....
-- 
				Valdis Kletnieks
				Operating Systems Analyst
				Virginia Tech

Attachment: pgpBSEz6ZTqjC.pgp
Description: PGP signature



[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]