g2k22 - OpenBSD hackathon Burg Liebenzell# 23 Sep 2022
Since I have been living near Stuttgart for over a year I decided to travel to g2k22 by bike.
The days before g2k22 I was a little unmotivated because I didn’t know what to work on. My yearly huge KDE Gear update to 22.08 was already committed. Even CMake 2.24.1 has already made it in.
When I arrived at the castle, I met sdk@ first. I asked him if he could build the Qt6 for me and at the and of the first day I got an OK from him and we have almost Qt6 imported in OpenBSD now. Of course without x11/qt6/qtwebengine I’ll hack on it if we see consumers otherwise wasting several hours of work makes no sense.
Next “big” thing on my list was
cmake.port.mk. I thought we could certainly
improve that. I end up with 3 changes:
cmake(1) controls the build and install task/jobs. This is the way CMake prefers and it works very well for us. Full bulk build passed! Victims already fixed.
Remove USE_NINJA from consumers Is no longer necessary but still optional (requested by sthen@). This diff removes all usage. Expect www/h2o (see at ports@) This is possible because of the use of cmake(1).
Improve Verbose mode
Before: ===> Regression tests for cmark-0.30.2p0 Test project /usr/ports/pobj/cmark-0.30.2/build-amd64 Start 1: api_test 1/9 Test #1: api_test ......................... Passed 0.01 sec Start 2: html_normalization 2/9 Test #2: html_normalization ............... Passed 0.16 sec Start 3: spectest_library
With the new option
MODCMAKE_VERBOSE (default to yes):
===> Regression tests for cmark-0.30.2p0 UpdateCTestConfiguration from :/usr/ports/pobj/cmark-0.30.2/build-amd64/DartConfiguration.tcl 21:09:28 [17/151] Test project /usr/ports/pobj/cmark-0.30.2/build-amd64 Constructing a list of tests Done constructing a list of tests Updating test list for fixtures Added 0 tests to meet fixture requirements Checking test dependency graph... Checking test dependency graph end test 1 Start 1: api_test 1: Test command: /usr/ports/pobj/cmark-0.30.2/build-amd64/api_test/api_test 1: Working Directory: /usr/ports/pobj/cmark-0.30.2/build-amd64/testdir 1: Test timeout computed to be: 10000000 1: 539 tests passed, 0 failed, 0 skipped 1: PASS 1/9 Test #1: api_test ......................... Passed 0.01 sec test 2 Start 2: html_normalization 2: Test command: /usr/local/bin/python3.9 "-m" "doctest" "/usr/ports/pobj/cmark-0.30.2/cmark-0.30.2/test/normalize.py" 2: Working Directory: /usr/ports/pobj/cmark-0.30.2/build-amd64/testdir 2: Test timeout computed to be: 10000000 2/9 Test #2: html_normalization ............... Passed 0.16 sec
That was easier than I thought and tb@ was kind enough to start an bulk build for me. Expect from 2-3 broken ports not much had been broken. I was able to fix the two victims quickly. I’m starting to be really happy with our CMake port and module.
Apart from that, here’s a short list of what I’ve been working on.
- devel/boost 1.80
- Update to 1.80 and backport some patches from upstram:
- Fix: Boost.Unordered containers are not in a valid state after moving
- qcad to 220.127.116.11
- gpgme to 1.18.0
- Remove mlt6 and webvfx
- Update quazip to 1.3
- chessx to 1.5.6
- krita to 5.1.0
- doxygen to 1.9.5
- cmark to 0.30.2
All in all I can look back on a really successful hackathon. Since I mainly operate only in ports, it was nice to meet people from the kernel space. On Monday morning I left the castle in the direction of Nürnberg where I worked for a customer for the next two days.