[development] RFC: drupal as a moving target

[catch]

I had my tongue slightly in my cheek while typing that, but still. The
intent behind that specific point was that I'd like to see golden contrib
(which exists in practice whether or not we call it that or have any visible
sign of it on Drupal.org) get a lot more issue queue attention than it
currently does. More testing and reviewing (and submitting) of patches, more
people tracking the lists to put down bogus/duplicate bug reports and answer
support requests, etc. etc. I have no illusions about this happening
immediately - there's a shortage of core patch reviews as it is, but it's
happened a bit with Views/CCK/Panels/Imagefield/Imagecache and there's a lot
of potential there for it to continue and develop. If there was a general
feeling (not a new rule) that we can't do a full release without x and y
modules, then some of the help that's ready to be  welcomed with open arms
might be more forthcoming. Additionally, the process of porting these
modules uncovers core bugs (or more often unforeseen deficiencies which are
harder to test for) every so often, and that would be happening during beta
rather than after .0

I don't have particularly strong views on this (except that it's worth
discussing), but, in the back of my mind there's a longer term idea for
somehow having HEAD ports of a few modules maintained, fully simpletested,
alongside core, so we can see exactly what breaks and how during the release
cycle as patches land (this long term idea stolen almost verbatim and
without permission from Douglas Hubler at the testing BOF). We have this to
an extent with Poll and Profile etc., but without simpletests it's been easy
for them to stay broken for quite a long time without anyone noticing - not
to mention they're not going to be exposing the same deficiences.


>
>
Steven Peck wrote:
> the best thing to do is
> use this time to get that desired functionality into core.

Well in this case the people best placed to do that are busy porting those
same modules to Drupal 6. I have high hopes for a longer code thaw, am
alpha/beta testing CCK and views ports, and will one of the first in line to
review any core patches that arrive in the queue, but for example, I can see
Panels and Nodequeue becoming as central to many sites and dependent contrib
modules in D6, as CCK and Views became in 5.x, so this is really more of a
general argument than about CCK and Views specifically (and sorry to pick on
Earl's modules but it's his own fault for coming up with them).
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.drupal.org/pipermail/development/attachments/20080430/deef3256/attachment-0001.htm