> Mike Meyer wrote:
>> On Fri, 16 Jan 2009 17:42:00 -0500 (EST) Dennis Clarke
>> <[email protected]> wrote:
>>> If the Solaris commands become a superset of the Gnu ones, then that
>>> position becomes a fait accompli.
>>
>> Thus avoiding the entire question of whether or not that's the best -
>> or even a desirable - goal.
>
> There are lots of proposals for someone else to do work, but few
> volunteers stepping up to the plate.

well hold on a sec there.

I was just looking at using a search for "gnu" in arg[0] in order to
determine if the current program was supposed to demonstrate GNUism or
UNIXism. The first thought that hits me is that a "grand unified" ls would
be needed ( and all other things that overlap in XPG4 land ) where each
executable must determine its own nature by both PATH and by options
provided. The entire idea is weak because one may move the executable
/usr/gnu/bin/foo to /tmp/foo and then a default behavior must be assumed.

I am just thinking out loud here.

If I go with some simplistic hack that checks if ( strstr( argv[0], "ucb"
) != NULL ) sort of thing then I can get away with this :

bash-3.2$ pwd
/export/home/dclarke/pgm/C/gnu
bash-3.2$ cc $CFLAGS -o ls hack.c
bash-3.2$ ./ls -l -v --debug
DEBUG: UNIXism assumed
ls: illegal option -- v
usage: ls -1RaAdCxmnlogrtucpFbqisfL [files]
bash-3.2$ cd ..
bash-3.2$ ./gnu/ls -l -v --debug
DEBUG: gnuism required
total 10
-rw-r--r-- 1 dclarke csw  683 Jan 16 22:28 hack.c
-rwxr-xr-x 1 dclarke csw 8512 Jan 16 22:29 ls
bash-3.2$ mkdir ucb
bash-3.2$ cp -p gnu/ls ucb
bash-3.2$ ./ucb/ls -l -v --debug
DEBUG : ucb behavior required
total 9
-rw-r--r--   1 dclarke       683 Jan 16 22:28 hack.c
-rwxr-xr-x   1 dclarke      7888 Jan 16 22:29 ls

Yes, this is messy.

It is not at all the sort of thing that happens in a single one nighter
and I don't even know if it is possible. Or reasonable. Regardless of the
need for someone to step up to the plate this requires some careful
thought first.

Dennis

_______________________________________________
opensolaris-discuss mailing list
[email protected]

Reply via email to