On Mon, Jun 28, 2004 at 09:03:29PM +0200, Ulf Lamping wrote:
> -buggy changes from tmpnam to mkstemp by Joerg Mayer doesn't seemed to
> be fixed (he promised to do so)
He backed them out.
The fix for the merge stuff is probably to:
have the "out_filename" argument to "merge_n_files()" be a "char
**", and, if the "char *" to which it points is null, have it mean
"open a temporary file" and return a pointer to that file through
"out_filename" (or have it return the filename as the return
value, with NULL meaning "error");
have "merge_open_outfile()" treat a null "out_file->filename" as
meaning that a temporary file should be used, use "mkstemp()" to
create the file, and use "wtap_dump_fdopen()" to open the file
for dumping.
The fix for the Windows printing stuff is probably to see if we can make
it work without a temporary file. I think that might be possible, but
I'd have to take a look at the code again to be sure.
> BTW: What's the state of installing the GTK-Wimp thing by the NSIS
> installer (Win32)?
The current state is that it lets you install with GTK-Wimp even if
that'd really really really really really really really really suck
because the standard Windows system font isn't scalable.
> BTW: Has the time come for a bug tracking system? Or maybe do we have
> one I just don't know of?
Well, we have a SourceForge project, and there are even some bugs there,
but it's probably not all that well known. (Tcpdump and libpcap also
have projects, and it's been a while since I looked at the archives for
"patches@xxxxxxxxxxx", because it's a truly hellish spam trap - unlike
SourceForge's bug tracking system.)
