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

Re: follow up on bs/del

On Tue, 14 May 1996, Carlos Carvalho wrote:

>  > 2. make fvwm run your script rather than netscape.  Edit your
>  > .fvwmrc and put in the full pathname to your netscape script.
> That's what I did.

Strange.  It should work then.

>  > #! /bin/bash
> I used sh, it's the same in debian.

True.  I have a habit of explicitly using /bin/bash when i write shell
scripts...i also use non-linux, non-gnu machines and sh can't always run
bash scripts.  bash isn't sh...it's an enhanced sh.

> My script does exactly the same; it puts the variables in the
> environment of the command only, not on the user environment.
>  > I use 'exec netscape' because the shell isn't needed any more after
>  > netscape loads...may as well get rid of immediately.
> Correct. However in this case you don't need to export the vars, do
> you?
>  >1.  Make sure the XNLSPATH and XKEYSYMDB env vars are set and exported
>  >BEFORE fvwm is run (i.e. before you startx).  This way, fvwm will inherit
>  >the environment vars from the parent shell, and everything that fvwm forks
>  >will inherit them too.
> As before, they're exported to the netscape process, which should be
> enough.

oh yeah.  I see what you were doing now.  Yep, that should work.

try my script and see if it makes any difference.  Maybe the variables DO
have to be explicitly exported.

It won't hurt to have the vars exported to your global environment
anyway...and will actually be of use if you use other motif apps.  Can't
see why anyone would want to unless they had bought a motif dev package
and could use shared libs rather than bloated static libs though...IMO
motif just isn't nice enough to be worth sacrificing dynamic libs for.

Also IMO, companies who make motif libs for linux should make the shared
libs (but not the headers) freely distributable under a "these may only
be used for pre-compiled binaries" license.  Then there'd be more linux
people willing to use motif apps, and more people would develop for
it...who wants to distribute a bloated static linked binary which is 2
or 4 times bigger than it needs to be?

Can't wait for the free lesstif clone... :-) Then I'll start exploring
motif in greater depth...until then I don't trust it because I have to
depend on somebody else to compile the binaries, which means I have to
trust them not to sneak in a trojan.


Reply to: