Hello. On 17/08/16 11:40, Stefan Schmidt wrote: > Hello. > > On 17/08/16 10:40, Stefan Schmidt wrote: >> Hello. >> >> On 17/08/16 01:14, Carsten Haitzler (The Rasterman) wrote: >>> On Wed, 17 Aug 2016 00:14:40 +0500 Nikita Zlobin <[email protected]> >>> said: >>> >>>> Hit, it is gentoo S). And this is not first case. >>>> But in turn, i remember case - when i took part in ladish development, >>>> i used ubuntu... and my code, being ok on my system, segfailed on >>>> gentoo system of project leader, revealing one hard to detect bug. >>>> >>>> doxygen 1.8.11, gcc 4.9.3 (though how it relates to doc gen...) >>>> >>>> As for configuration, i took them from ebuild log: >>>> $ ./configure --enable-cserve --enable-image-loader-generic >>>> --enable-image-loader-jpeg --disable-tizen --disable-gesture >>>> --disable-gstreamer --enable-xinput2 --disable-xinput22 >>>> --disable-multisense --enable-libmount --with-opengl-full >>>> --disable-physics --with-glib=no --enable-image-loader-jp2k >>>> --enable-gnutls >>>> --enable-i-really-know-what-i-am-doing-and-that-this-will-probably-break-things-and-i-will-fix-them-myself-and-send-patches-abb >>>> --prefix=/usr >>>> >>>> As for bugs - you can see there one veeery looooonnng option, so >>>> decided to consult. Can't beleive, that these options might cause doc >>>> making to fail. >>> >>> --enable-cserve <- you know this is enabled by default? why enable? >>> --enable-image-loader-generic <- enabled by default too >>> --enable---disable-image-loader-jpeg <- enabled by default too >>> --disable-tizen <- disabled by defaault... >>> --disable-gesture <- disabled by default... >>> --disable-gstreamer <- you know this is disabled by default... >>> --disable-multisense <- you do know this breaks theme functionality right? >>> --enable-libmount <- you know itss enabled by default? >>> --with-opengl-full <- you know this is a typo? it should be: >>> --with-opengl=full ... and this is the default anyway... >>> --disable-physics <- you know this can break theme functionality right? >>> --with-glib=no <- you know this can create problems with things like the >>> ibus >>> imf modules which require it (because libibus need glib mainloop stuff), >>> though you don't enable the ibus imf module here...? >>> --enable-gnutls <- why gnutls not openssl. you know we test openssl as it's >>> the default. gnutls gets little to no testing. also optimization - little >>> to >>> none with it. >>> >>> almost all your options are redundant and several i would think are dubious >>> to >>> alter... though none of these SHOULD affect document generation - maybe >>> unless >>> there is more glib main loop integration that is needed that we've missed >>> (scim? something else?) and since it's on by default ... we haven't seen the >>> brokenness with it off. there is a reason we make you put in that insanely >>> long >>> option accepting your fate if you stray from tested defaults that might >>> have an >>> effect. >> >> We can easily tell if it works with the default configure options on his >> system. Nikita, can you run a configure without these options (any >> options at all) followed by a make and a make doc? >> >> This is not going to install anything so it will not touch your system. >> >> If that still breaks the same way as before it is more likely a problem >> with your doxygen version. It works for me here with 1.8.9.1 and yours >> is 1.8.11. So if it still fails the next try might be to downgrade your >> doxygen version and see if that helps. > > I did the inverted test and tried with doxygen 1.8.11 (which is the > latest release, December 2015) and can verify that it breaks our doc > build the way you reported. > > Latest doxygen from their git repo works though. So it is really a > problem in 1.8.11 (maybe also 1.8.10) and is fixed already. We can only > hope for a new release or go back to an older version.
Doxygen 1.8.12 was released a few days ago and I did a local test with it. Works fine. regards Stefan Schmidt ------------------------------------------------------------------------------ _______________________________________________ enlightenment-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
