From #gnunet: <dduffield> Hello, I am a Gentoo user, and I have a build failure for 0.9.5a, where ld returned 1 exit status, involving undefiend reference to 'MHD_create_post_processor' in gnunet-gns-fcfsd.c <grothoff> dduffield: yes, somehow Gentoo must have build with --disable-postprocessor or something like that, which is explicitly something distributions must NOT do. (This is an option for embedded systems where 1k of code size matters.) Checking the ebuilds, there's a USE="messages" that controls the postprocessor feature indirectly. This might not be an intended feature, and needlessly confuses users. Maybe it should be uncoupled from USE="messages" and enabled unconditionally?
(In reply to Patrick Lauer from comment #0) > From #gnunet: > > <dduffield> Hello, I am a Gentoo user, and I have a build failure for > 0.9.5a, where ld returned 1 exit status, involving undefiend reference to > 'MHD_create_post_processor' in gnunet-gns-fcfsd.c > <grothoff> dduffield: yes, somehow Gentoo must have build with > --disable-postprocessor or something like that, which is explicitly > something distributions must NOT do. (This is an option for embedded systems > where 1k of code size matters.) > > Checking the ebuilds, there's a USE="messages" that controls the > postprocessor feature indirectly. > > This might not be an intended feature, and needlessly confuses users. > Maybe it should be uncoupled from USE="messages" and enabled unconditionally? Decoupling it from messages works. Since it failed otherwise, no rev bump needed. Its in the tree for 0.9.42, previous unstable versions are off. I'm opening a stablereq for 0.9.42 since its been in long enough.