[development] Drupal and acceptance testing

Caleb Gilbert caleb.gilbert at gmail.com
Mon Aug 25 18:43:56 UTC 2008


Thanks for the reply, Catch - good stuff. I'm going to take a look at what
it would take to get Selenium RC to work with simpletest instead of PHPunit
(the unit testing framework the Selenium RC docs refer to).

An interesting note about Selenium's ties to Drupal - the new AutoPilot 2.0
beta - which was just released in time for Drupalicon -
http://twitter.com/workhabit  heavily relies on  selenium (e.g., selenium's
macro recorder/playback is being used in this case for handling change
management [pushing changes from dev to staging to production], so
apparently it is not just for acceptance testing!).

- Caleb


I suggested an automated javascript testing framework for gsoc, was thinking
watir rather than selenium based on a few discussions about both, but it
didn't go anywhere:http://groups.drupal.org/node/9511,

While I'd love to see testing with Selenium or Watir or some other browser
remote control that can handle stuff that simpletest or straight up
javascript unit testing can't, like cwgordon said, it's a matter of effort
and returns.

The testing framework we have in core has the advantage that anyone can
download and run tests without needing to install other software, tests can
be written in PHP, and we're getting close to the tipping point where our
coverage is high enough to start finding regressions before they're
committed.

On the other hand, and  unfortunately, at the moment we don't have that many
jQuery intensive patches - yes, it's really annoying trying to get
cross-browser tests when a new jQuery is released, but that's once every few
months - so in terms of saving click around time SimpleTest is giving higher
returns at the moment. While it might appear to be a lack of interest at the
moment, it's more a case of 1. a lack of resources (the 'testing brigade' is
still pretty small) 2. a need to get the testing framework we actually have
up to a point where it begins to actually save some work - it'll only do
this when we're closer to 100% test coverage and tests are being run on
patches automatically via testing.drupal.org

That said - it's very clear that the core testing framework can't cover
everything, so extending it would be the natural next step - and given the
amount of time and work it's taken to get to where we are now, starting
early surely wouldn't hurt.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.drupal.org/pipermail/development/attachments/20080825/ea46f2eb/attachment.htm 


More information about the development mailing list