Re: weird PATH_MAX discussion in mplayer's list. (Fwd: [MPlayer-dev-eng] [PATCH] fix for unconditional use of PATH_MAX)

[Marcus Brinkmann]

On Thu, May 01, 2003 at 07:15:38PM +0200, Robert Millan wrote:
> please someone could lend me a hand? i'm having trouble with these
> people trying to convince them to support systems without PATH_MAX

The reason not to have a fixed limit is not to allow people to use
500.000.000 character long path names, but to have path names that are 513
characters long (for example).

However, it seems to be useless to argue with these people.  They know the
problem, they know that they are not writing software compatible to all
POSIX systems, and they know the proper way to fix it if they want to.
Note that the arguments against the fix are non-sense:  Neither is a loop
over realloc complicated (or they should stay away from writing software at
all!), nor is enforcing resource limits a job of mplayer, but a job of the
operating system, and the mplayer people should stick with designing and
writing a good media player, and leave the job to decide what system calls
with which parameters are valid and make sense to people designing and
writing operating systems.

If they choose to deliberately write a bug into their software, then that is
their responsibility.  If they hard code an

#ifndef PATH_MAX
#define PATH_MAX 512
#endif

then their software will compile just fine on the Hurd and run, that is
until people suddenly start to use it with path names longer than that, of
course.  I would recommend 4096, though, that buys you a couple of more
years time until the problem arises.

"Certainly everybody will have replaced this program by the year 2000."

Thanks,
Marcus

-- 
`Rhubarb is no Egyptian god.' GNU      http://www.gnu.org    marcus@gnu.org
Marcus Brinkmann              The Hurd http://www.gnu.org/software/hurd/
Marcus.Brinkmann@ruhr-uni-bochum.de
http://www.marcus-brinkmann.de/