| Re: [AD] [PATCH] Solaris Fixes and dev notes |
[ Thread Index |
Date Index
| More lists.liballeg.org/allegro-developers Archives
]
On Sunday 05 June 2005 19:13, Shawn Walker wrote:
> > Rather than checking for any of the sun preprocessor directives
> > (which are unreliable, as I recall) we should have the configure script
> > check if the function exists - afterall, we don't want to know if we're
on
> > Solaris, but if getexecname() works!
>
>
> That's sounds better certainly, but I don't know how to do that :) I
don't
> use configure or the like in any of my personal projects.
That's ok, I usually look at the existing configure.in and aclocal.m4 to
get an idea on how to do things myself and then read the manual if I can't
figure it out.
> It also seems to me that the BSD's might use this same code section (the
> System V way)?
Yes, but without the argv parameter.
> Well, I had already covered the concern of the function not suceeding.
But,
> it looks like the only additional possibility is that sometimes it might
> return a relative path to the executable instead of an absolute path. It
> suggests pre-pending the output of getcwd in such a case. Otherwise, I
think
> it's definitely the preferred way to go about this. I can make the
necessary
> change to have it account for the relative path case, as I'm not
concerned
> about the warning regarding getcwd not working: "unless the process or
one
> of its ancestors has changed its root directory or current working
directory
> since the last successful call to one of the exec family of functions".
> Since this happens at the beginning of the allegro init cycle. Once
someone
> changes the current working directory all bets are off anyway.
The problem is, someone might actually do just that. It just occured to me
that the return value of get_executable_name() would not change during the
execution of the programme, so redoing all the work all the time seems a
bit silly. I'll see if I can address both issues.
> However you folks want to work it out, I just can't use it right now
unless
> I make those changes.
Right, I'm not saying it shouldn't be fixed, just fixed in a different way.
Evert