On Sun 2016-01-03 08:32:39 -0500, Patrick Brunschwig wrote: > > I have created the first *beta* release of Enigmail v1.9.
Thanks, Patrick, this is great to have! It already warned me about the fact that my OpenPGP certificate is going to expire in less than 30 days. And i'm quite happy about dropping the binary parts of the build. Kudos! I've just uploaded 1.9~beta1-1 to debian experimental, so it should be available in a few hours for anyone running from that repo. A handful of nit-picky notes from my inspection of the build: A) package signatures ------------------ > I would appreciate to get as many bugs reported as possible such that > we will have a stable release by the end of February. > > The package (XPI) is available from > https://www.enigmail.net/download/beta/enigmail-1.9-beta1.xpi > > SHA-1: c9b4b2e943134549b07ceffd880605e4b98ca1a5 I recommend using SHA-256 checksums for binary artifacts like zipfiles these days :) Do you have plans to release a signed "source tarball" for this as well? I'm happy to work from the signed git tags if you prefer, but sometimes it's nice to have a cryptographically-signed tarball with detached signature that can be used directly (e.g. in the debian archive). B) unused properties ----------------- running "make check" shows, among some good output, some "false positives": ============================================= dtdLabels: 613 found Labels: 610 false positive (or correct because in comment)?: enigmail.msgViewColumn.tooltip unused labels in 'unused.txt' ============================================= propLabels: 474 found Props: 446 false positive (or correct because in comment)?: clickDecrypt false positive (or correct because in comment)?: cmdLine false positive (or correct because in comment)?: debugLog.title false positive (or correct because in comment)?: encryptYes false positive (or correct because in comment)?: internalError false positive (or correct because in comment)?: keyTypePrimary false positive (or correct because in comment)?: msgCompose.button.save false positive (or correct because in comment)?: pubKey false positive (or correct because in comment)?: sendKeysOk false positive (or correct because in comment)?: signOff false positive (or correct because in comment)?: signOn false positive (or correct because in comment)?: signYes false positive (or correct because in comment)?: warning and unused.txt contains: ++++++++++++ unused labels: unused properties: addUidDlg.commentError errorKeyUnusable gnupg.invalidKey.desc ++++++++++++ C) make test and test.sh --------------------- "make check" and "make unit" are now subdependencies of an otherwise empty "make test". However, at least with my version of GNU make (4.0), the existence of "test.sh" appears to imply an implicit rule that does this: cat test.sh >test chmod a+x test so i end up with an extra file named "test" that is an exact copy of test.sh. Not sure what the right way to clean this up is, but it makes my build slightly messier than it needs to be. D) jshint ------ The test suite now uses jshint, but there's no mention of jshint in ./COMPILING -- maybe it's worth documenting the test process in COMPILING as well? Also, I hate to bug you about this, but jshint isn't licensed as free software because of crockford's "good not evil" clause. This means that debian won't be able to run this particular part of the tests in its autobuild setups for you. see https://github.com/jshint/jshint/issues/1234 for ongoing (multiyear!) work there :/ sigh. If you can use a different linter (or want to prod jshint upstream about continuing work to fix this), i can get those tests running on the debian infrastructure as well as anyone who builds debian packages from source too. All the best, --dkg _______________________________________________ enigmail-users mailing list [email protected] To unsubscribe or make changes to your subscription click here: https://admin.hostpoint.ch/mailman/listinfo/enigmail-users_enigmail.net
