Created attachment 168625 [details] Patch to be applied to www/uzbl to update to version 0.9. Update uzbl to Version 0.9. For release notes see http://www.uzbl.org/news.php?id=41 Note that the syntax of the configuration file has changed in an incompatible way. While there, also change port to download tar balls directly from github.
Note that the patch - removes files/patch-examples__data__scripts__auth.py as this change is already upstream - add a new file files/patch-src__requests.c
Can you test the port with portlint -AC ? There are several issues that can be improved. For example: the man pages are not installed gzipped.
Created attachment 168924 [details] Patch to be applied to www/uzbl to update to version 0.9.
Patch amended according to the suggestions of portlint -AC.
testbuilds@work
The tests take really long, it builds www/webkit-gtk2 as dependency, which really takes very long. The first build on current (11a) failed, expecting python3 somewhere, see http://people.freebsd.org/~pi/logs/www__uzbl-cur-1459695703.txt The other ones will probably fail in a similar fashion, still running. Can you investigate ?
The problem is the MAKE_ENV; somehow, in your test build, my addition MAKE_ENV+= PYTHON=${PYTHON_VERSION} to the ports Makefile got lost or otherwise took no effect. I'll try to find out how this can happen, but I'll have a hard time to reproduce, given that at my machine (stable/10 with ports from head) make -V MAKE_ENV shows the correct environment starting with PYTHON=python2.7 and hence everything works correctly with using only python2.
Please note that I build with poudriere -- have you tried it with poudriere, too ?
Created attachment 168952 [details] Patch to be applied to www/uzbl to update to version 0.9.
I haven't tried it with poudire; however, since poudire also generates the --MAKE_ENV-- debug output with make -V, it was clear that the Makefile was wrong. Looking at the id-strings, I could also reproduce what went wrong: www/uzbl was updated after(!) I originally submitted this PR, so my patches were still patches against the port as it was on the original submission date. That caused the hunk containing the "MAKE_ENV+= PYTHON=${PYTHON_VERSION}" to not fail. I've refreshed the patch; can you please try the refreshed patch?
New patch does no longer show the build problem that Kurt has identified with the previous patch.
A commit references this bug: Author: riggs Date: Sun Apr 10 12:17:05 UTC 2016 New revision: 412925 URL: https://svnweb.freebsd.org/changeset/ports/412925 Log: Update to upstream version 0.9 PR: 208293 Submitted by: aehlig@linta.de (maintainer) Changes: head/www/uzbl/Makefile head/www/uzbl/distinfo head/www/uzbl/files/patch-Makefile head/www/uzbl/files/patch-examples__data__scripts__auth.py head/www/uzbl/files/patch-examples__data__scripts__instance-select-wii.sh head/www/uzbl/files/patch-examples__data__scripts__load_url_from_history.sh head/www/uzbl/files/patch-src__events.c head/www/uzbl/files/patch-src__requests.c head/www/uzbl/files/patch-src__uzbl-browser head/www/uzbl/files/patch-uzbl-core.c head/www/uzbl/pkg-plist