Surely almost all Sage releases contain major new functionality! Except very occasionally a release is followed by a quick emergency bug fix. And also rather occasionally there is a huge new feature.
Thus I would expect that * most releases sage-x.y.z would increment y and reset z (to 0, or 1). * small quick emergency bugfix releases would just increment z * really major releases (e.g. 100% doctests! or the Windows port!) would increment x (and reset y and z to 0, or 1). This is close to what we do except that z is incremented when it perhaps ought to have been y. For example, 4.3.2 should perhaps have been 4.4, while 5.*.* is waiting for a really big milestone (which we should perhaps decide on in advance, as a big but achievable target). For a start, please let's decide on what's to come next after 4.3.2. Today I checked in a patch (including a bugfixed spkg) and had to mark it 4.3.2 since there was nothing else! Of course it should not go into 4.3.2 since that's now at rc0, hence on feature-freeze. John On 4 February 2010 16:38, Dr. David Kirkby <david.kir...@onetel.net> wrote: > Harald Schilly wrote: > >> ...and to come back on topic: looking back, i think the best version >> number for Sage were the release dates (back in the early 0.x time). >> Today, a "YYYY-MM" string would be nice (adding a .1 if there is a >> second release in the same month). >> >> H > > I can't say I agree with that. > > Even though the Y.Z numbers of X.Y.Z don't seem to mean a lot, I assume if X > was updated, it would be believed there were significant changes. With a > date, one knows very little indeed about the amount of change. (Of course, > you know more about the date). > > Rather than add a .1 if there was two releases in the same month, would it > not be simpler to just have the release date in full? But I still think the > current numbering is better than a date, even though I have my concerns > about the current numbering method. > > Dave > > > -- > To post to this group, send an email to sage-devel@googlegroups.com > To unsubscribe from this group, send an email to > sage-devel+unsubscr...@googlegroups.com > For more options, visit this group at > http://groups.google.com/group/sage-devel > URL: http://www.sagemath.org > -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
