[QGIS-Developer] Realistically, we can't release in 2 days ?

[Greg Troxel]

Jürgen E. Fischer via QGIS-Developer <qgis-developer at lists.osgeo.org>
writes:

> If we have a blocker we should postpone.  How much time do we need?

>From my viewpoint, there need to be an rc for a while before a 4.0, so
it's not like waiting 5 days could address it.  Realistically I cannot
see a 4.0 (which has not been unreasonably rushed) before 2 weeks from
now, even if a fix for the last problems magically lands today.

> I suppose we're just moving the release date ahead and don't do additional
> releases of 3.40 and 3.44 instead?

>From my packager viewpoint, I would prefer one of

  - micro releases of 3.40 and 3.44 on schedule, because there's no
    reason not to.  I really don't care that the eventual 4.0 release is
    or isn't synchronized to 3.4*.x point releases.  Updating to new
    point releases is really easy - just change the version, regen
    checksum, build/replace package and start up qgis on a project, push
    changes.  It's literally less than 10 minutes of effort and I have
    no memory of it going badly.

  - as above, but when releassing 3.40.x and 3.44.y, decide that 3.44 is
    now LTR.   The fact that 4.0 is delayed is not related to "3.44 is
    fine and is what people ought to be running until their packaging
    system and they decide that it's wise to run 4.0".


(I don't personally care which one has the LTR label as I have already
changed pkgsrc, which has historically been LTRs, to be 3.44 a few
months ago.)