[Grok-dev] Grok 1.1 important issues

[Jan-Wijbrand Kolman]

Hi Steve,

Steve Schmechel <steveschmechel at yahoo.com> wrote: 
> I would like to help test this on Windows with Python 2.6, but I want
> to be sure that my procedure is correct for what you intend to
> determine.
> 
> Could you please describe (or point to a document that describes) the
> procedure for testing a potential Grok release?  
> 
> Are you testing the latest trunk code (checkout and run buildout), or
> are you starting with grokproject and Grok version 1.1a2 and then
> pulling in newer code?
> 
> Are there any specific changes to the buildout.cfg or other
> configuration files?
> 
> How deep does the testing go?  Do we just run Grok-specific tests or
> recurse into all the tests in the underlying packages and libraries?
> 
> How many tests are expected to run and "pass"?
> 
> I just want to be sure that if I say "the tests pass", we are talking
> about the same tests.  (Assumptions are more dangerous if I do not
> find a problem, because there will be no need for someone to further
> investigate some specific finding.)

Good questions. Let's see if I can answer them.

In all cases I relied on the test coverage of the various packages 
involved.

I did a checkout of the groktoolkit::

  svn co http://svn.zope.org/repos/main/groktoolkit/trunk.

With the python version I wanted to test with, I ran the bootstrap.py 
for the groktoolkit checkout. Then run ./bin/buildout. Then you can run 
the tests for grok and the grokcore.* family and martian - each in 
eachother's context::

  ./bin/test-grok

Since the groktoolkit points to the latest 1.1a2 release, you're not yet 
testing the grok trunk itself. To run all of these test against the grok 
trunk you can use the mr.developer extension script::

  ./bin/develop activate grok$

(Note the trailing "$". This is needed because the develop command will 
regard the name as a regex and without it, all packages with grok in 
their name are checked out).

I then ran buildout again to have the script use this grok trunk 
checkout::

  ./bin/buildout

and ran the tests again::

  ./bin/test-grok

There should be no buildout errors nor test failures whatsoever.


After that I reset the buildout::

  ./bin/develop reset

  ./bin/buildout

The I removed the .installed.cfg, bin directory , parts directory and 
develop-eggs directory before re-running the bootstrap.py script with a 
different python version. Using a completely fresh groktoolkit checkout 
would work to of course.

I hope this helps. I'm curious whether others would have any comments or 
suggestions to this approach.

regards,
jw