X-Git-Url: https://pere.pagekite.me/gitweb/homepage.git/blobdiff_plain/e80fd8769bcfb907ce5fa2597022b496eb658342..ae5db6d19f3d85fdd5e7bd4c12be28fa3f15fc43:/blog/archive/2014/04/04.rss diff --git a/blog/archive/2014/04/04.rss b/blog/archive/2014/04/04.rss index 9ae8ecaa23..71971f80f2 100644 --- a/blog/archive/2014/04/04.rss +++ b/blog/archive/2014/04/04.rss @@ -6,6 +6,55 @@ http://people.skolelinux.org/pere/blog/ + + Half the Coverity issues in Gnash fixed in the next release + http://people.skolelinux.org/pere/blog/Half_the_Coverity_issues_in_Gnash_fixed_in_the_next_release.html + http://people.skolelinux.org/pere/blog/Half_the_Coverity_issues_in_Gnash_fixed_in_the_next_release.html + Tue, 29 Apr 2014 14:20:00 +0200 + <p>I've been following <a href="http://www.getgnash.org/">the Gnash +project</a> for quite a while now. It is a free software +implementation of Adobe Flash, both a standalone player and a browser +plugin. Gnash implement support for the AVM1 format (and not the +newer AVM2 format - see +<a href="http://lightspark.github.io/">Lightspark</a> for that one), +allowing several flash based sites to work. Thanks to the friendly +developers at Youtube, it also work with Youtube videos, because the +Javascript code at Youtube detect Gnash and serve a AVM1 player to +those users. :) Would be great if someone found time to implement AVM2 +support, but it has not happened yet. If you install both Lightspark +and Gnash, Lightspark will invoke Gnash if it find a AVM1 flash file, +so you can get both handled as free software. Unfortunately, +Lightspark so far only implement a small subset of AVM2, and many +sites do not work yet.</p> + +<p>A few months ago, I started looking at +<a href="http://scan.coverity.com/">Coverity</a>, the static source +checker used to find heaps and heaps of bugs in free software (thanks +to the donation of a scanning service to free software projects by the +company developing this non-free code checker), and Gnash was one of +the projects I decided to check out. Coverity is able to find lock +errors, memory errors, dead code and more. A few days ago they even +extended it to also be able to find the heartbleed bug in OpenSSL. +There are heaps of checks being done on the instrumented code, and the +amount of bogus warnings is quite low compared to the other static +code checkers I have tested over the years.</p> + +<p>Since a few weeks ago, I've been working with the other Gnash +developers squashing bugs discovered by Coverity. I was quite happy +today when I checked the current status and saw that of the 777 issues +detected so far, 374 are marked as fixed. This make me confident that +the next Gnash release will be more stable and more dependable than +the previous one. Most of the reported issues were and are in the +test suite, but it also found a few in the rest of the code.</p> + +<p>If you want to help out, you find us on +<a href="https://lists.gnu.org/mailman/listinfo/gnash-dev">the +gnash-dev mailing list</a> and on +<a href="irc://irc.freenode.net/#gnash">the #gnash channel on +irc.freenode.net IRC server</a>.</p> + + + Install hardware dependent packages using tasksel (Isenkram 0.7) http://people.skolelinux.org/pere/blog/Install_hardware_dependent_packages_using_tasksel__Isenkram_0_7_.html @@ -69,7 +118,7 @@ the python-apt code (bug <a href="http://bugs.debian.org/745487">#745487</a>), but using a workaround I was able to get rid of the file descriptor leak and reduce the memory leak from ~30 MiB per hardware detection down to -around 2 miB per hardware detection. It should make the desktop +around 2 MiB per hardware detection. It should make the desktop daemon a lot more useful. The fix is in version 0.7 uploaded to unstable today.</p>