[Framework-Team] Re: PLIP #234 Review Revisions
optilude+lists at gmail.com
Thu Feb 12 13:40:57 UTC 2009
Calvin Hendryx-Parker wrote:
> I'd be open for suggestions for other tests. A majority of the fixes
> were templates and they were to use an already existing bit of code
> functionality that I didn't add like you said. The viewlets were
> modified to support this and I have the test to confirm it in the base
> class for the viewlets.
> If there are some specific tests you'd like to see I'd be happy to add
> them, but in the end the trickiest piece of code I updated was the
> calendar tool modifications and I added tests there also.
> FYI... we have released collective.lineage which depends on many of
> these fixes and I believe will expand the number of people using this
> functionality. There are a lot of people excited about this new
> product and are wanting to deploy it. We actually have 1 site in
> production with it already. Since this functionality was already in
> Plone before I got ahold of it I still see this PLIP as a large bug
> fix that really should be included.
FWIW, I think much of Calvin's work could've gone into a 3.2.x release
as bug fixes. If he doesn't break tests, and writes a few tests for
truly new code, then I think that's probably sufficient in most places.
Having done something similar in the past (but not read Calvin's diff in
detail), I suspect most of his changes were simply to stop people making
use of portal_url() when they should've used get_navigation_root().
There may be cases when we could add a "defect" type test to show that
the navigation root didn't work before, but now does, but let's not
create too much work for what is, in many cases, more about analysing a
problem and applying a few surgical fixes, than writing a ton of new code.
Author of `Professional Plone Development`, a book for developers who
want to work with Plone. See http://martinaspeli.net/plone-book
More information about the Framework-Team