Summary: | dbus-1.4 review wishlist | ||
---|---|---|---|
Product: | dbus | Reporter: | Simon McVittie <smcv> |
Component: | core | Assignee: | Havoc Pennington <hp> |
Status: | RESOLVED FIXED | QA Contact: | John (J5) Palmieri <johnp> |
Severity: | enhancement | ||
Priority: | high | CC: | cosimo.alfarano, hp, lennart, walters, will, zeuthen |
Version: | 1.4.x | ||
Hardware: | Other | ||
OS: | All | ||
Whiteboard: | |||
i915 platform: | i915 features: | ||
Bug Depends on: | 11191, 11668, 14512, 16621, 19681, 19997, 29228, 32827, 33465, 33466, 34405, 34569, 35173, 35228, 35229, 35358, 35705, 36130, 36156, 37258, 37847, 37870, 38005, 38120, 38656, 39987, 40003, 40004, 40235, 41047, 41218, 41219, 42811 | ||
Bug Blocks: | 36164 |
Description
Simon McVittie
2011-04-08 04:20:17 UTC
smcv's wishlist: * Bug #14512: build system badgering (install dbus-daemon less strangely) * Bug #16621: don't force -fPIE, it makes some people's toolchains explode and distributions that want PIE can get it just as easily without our help * Bug #19997: make regression tests, dbus-launch --exit-with-session work if X11 autolaunching is disabled * Bug #32827: use Libs.private in .pc file, add -uninstalled version * Bug #33465: make Solaris fd-passing work (dept. of portability) * Bug #33466: support --disable-gc-sections for broken toolchains * Bug #34569: embarrassingly, nonce-tcp transport has never worked on Unix * Bug #35173: fd leaks falsely reported if fd >= 3 was already open walters' wishlist (guessed): * Bug #35705: system-bus-activation logging nearly there Lennart's wishlist (guessed): * Bug #35228: use /etc/machine-id if available r+ modulo whitespace nitpicking. * Bug #35229: add /lib/dbus-1/system-services to the search path r- but I suggested changes that I'd accept Bug #35358, syslog when dropping messages, would be good for 1.4 too. Revisiting this list, in suggested review order: * Bug #33465 is trivial and helps portability to Solaris * Bug #29228 is trivial, fixes a rarely used setup (cmake on Unix), but makes it easier for me to test the cmake build system (which is popular on Windows) * Bug #37258 is a trivial fix for a fd leak * Bug #34405 is relatively simple build-system badgering that improves the "write test / hack until it passes" cycle * Bug #36156 is relatively simple build-system badgering that improves distro packages :-) * Bug #38005 needs more careful review, but the bugs it fixes would be incredibly hard to debug if they happened * Bug #35229 is review-, although I might hijack it and write a patch for it myself at some point Returning to this: Bug #11191 finally turns an internal union representing "thing that can be sent on D-Bus" (which more or less every binding currently has to reinvent) into public API. It could go to 1.4 or 1.5, I don't really care, but I'm tired of reimplementing it repeatedly. Bug #42811 improves test coverage in builds that don't have the extra test goo embedded in the library (which is documented to be a bad idea in "production" builds). |
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.