-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Guillaume Melquiond schrieb: > On 2009/3/17, Nils Kneuper wrote: > >> I just wanted to inform you about the deadline for commits that are meant to >> go >> into 1.6. That is everything in until tomorrow in the afternoon in Europe >> will >> be in, everything later will not. The firm "pencils down date" is as follows: >> >> March 18th, 5PM GMT+1 >> (as in "five in the afternoon in Europe (Paris/Berlin/...)") >> >> Currently I even think about placing an "svn lock" at this time to make sure >> that it is done for the moment. The lock will be released with a) tagging 1.6 >> and b) branching trunk over to branches/1.6. If you have *anything* else >> that I >> have to wait for, ping me early in IRC, so that I can consider pausing for >> your >> change/fix. If I get nothing, I will "just" get it done... > > I just ran a diffstat and this is what I obtain: > - for WML (data/): 86 files changed, 542 insertions(+), 566 deletions(-) > - for C++ (src/): 23 files changed, 171 insertions(+), 140 deletions(-) > > In other words, the 1.6 release will have hardly been tested by rc3. > So I would like to suggest yet another release candidate, so that the > 1.6 release actually gets some testing done. Ideally (and this also > happens to be true in most projects), the release should be identical > to the last release candidate, except for the version number.
You know, in a perfect world this would easily be possible. But in reality we see that after such a "release candidate that will just be relabeled" some "smaller commits" fixing smaller annoyances come in. Should those "tiny fixes" stay out if we repackage anyway? And if they should go in, which ones are "small enough" to be allowed and which ones are not? This is a dilemma we got and basically I stay with the opinion that current trunk is in a good enough shape to get it out as 1.6. Regarding the "I have a bad feeling" part: Yeah, I think we all got this feeling simply because the number of testers for 1.5.x was too low. But we won't really get more testers unless we label it "done". That is sadly the way it is. And regarding "we should wait more until the AI testsuite is ready": Which gain will we directly have? We will see how/if the AI changes behavior when altering parameters. We already know/have the suspicion that some AI params have close to no or even no effect. What else can we learn from it? We will have a testbed for further improvements. Those will take *time*. One week is clearly not enough for those. But we can still release 1.6.1 in lets say 3 weeks with some *real* fixes for this stuff. Personally I think that what we currently have in trunk is better than 1.4.x. Better regarding several "small" and many *big* annoyances (like the addon download!). And for a feeling "there might be some problems left" I won't hold back. We have no way to find the issues unless they get reported and this most likely won't happen unless we label 1.6 "done". In general we still got time till the announcement it published to see if there are any "real showstoppers" left. We can then still hammer out a hotfix release. I won't announce till Sunday, so this should leave us with enough time if anything *really* problematic occurs. So unless someone has a blocker right now, I will release in 5:30... Cheers, Nils Kneuper aka Ivanovic -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.10 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iEYEARECAAYFAknAzesACgkQfFda9thizwV4pgCgg3x3YJy7daQ7E+dN7kumJfru Ax8AoKHvMUHkgimXUD7gyYFKL96oUi3T =Aodx -----END PGP SIGNATURE----- _______________________________________________ Wesnoth-dev mailing list [email protected] https://mail.gna.org/listinfo/wesnoth-dev
