Wireshark-users: Re: [Wireshark-users] build problem
Date Index · Thread Index · Other Months · All Mailing Lists
Date Prev · Date Next · Thread Prev · Thread Next
From: Jeff Morriss <jeff.morriss.ws@xxxxxxxxx>
Date: Thu, 28 Feb 2008 16:59:20 -0500


Michael Tuexen wrote:

On Feb 28, 2008, at 4:30 PM, Jeff Morriss wrote:


Guy Harris wrote:

bitmus DA wrote:


i waited and downloaded version 0.99.8
then configured it --without-pcap and compiled. but error still here
That's another bug, not fixed in 0.99.8, but fixed in the current SVN, 
so that *particular* fix should be in the next release.

For better or worse, most developers don't build without libpcap, and
the buildbots don't build without libpcap, so build problems when
building without libpcap don't tend to get discovered and fixed.
Obviously, we don't want all the buildbots building *only* without libpcap.
We could perhaps do all builds first with, and then without, libpcap - or first without, and then with, libpcap, with post-build tests done on the versions with libpcap - although that would slow the buildbots down 
somewhat.

The Solaris buildbot appears to have taken about 4 3/4 minutes to run
the configure script, and about 1 hour 38 minutes to do the build,
unless I'm missing something, so it'd add about 1 3/4 hours to the
Solaris build procedure to do both builds.

However, if we *only* do it on, say, the OS X buildbot, it shouldn't
slow things down too much (about 15 minutes), and would probably catch 
most if not all of the problems with non-pcap builds.
Funny, I was going to suggest that one of the buildbots build with both GTK2 and GTK1 1 for the same reason. I agree that only one need do it, 
though.


So we need with-libpcap and without, with gtk1 or gtk2?
So the Mac OS X buildbot could build
- gtk2 without libpcap
- gtk1 with libpap
(I'm not sure if gtk1 without libpcap makes much sense)...


Yeah, that would cover most of the issues I think.