I would like to start a discussion about the future of this module. Since Simpletest has been moved into core the development of the contributed module has practically ground to a halt. This is very much understandable, having Simpletest in core is a very good thing. Ideally there should not be a need for this module anymore, and we should try to avoid any duplicate efforts in maintaining Simpletest both in core and contrib.

At the moment there are still some use cases that warrant the use of the contributed module. Not all functionality of the contributed module has been ported to core (for example DrupalCloneTestCase and DrupalRemoteTestCase). But at the same time the contrib module is starting to lag far behind in features and bugfixes, and realistically the chances that someone will step up and take over maintainership is rather low - Simpletest in core is much more attractive to work on, and I don't think anyone looks forward to having the job of backporting dozens of core issues.

A big problem is that both modules are mutually exclusive. You need Switcheroo to choose one or the other. This is not a sustainable situation.

So what can we do?

  1. Keep the contributed module but refactor it so it depends on core Simpletest? We could rename it Simpletest Extras, remove all code that is in core now and rework the additional classes so that it is compatible with core Simpletest again.
  2. Abandon contrib Simpletest for D7 and onwards, and make an effort to port all missing functionality to Simpletest core. Some of the functionality in contrib Simpletest needs a lot of work to meet the quality standards of core though.
  3. Weather the storm and do nothing since it is possible that Drupal will move to a different testing framework in the future, as per #1567500: [meta] Pave the way to replace the testing framework with PHPUnit and possibly rewrite Simpletest.

Thoughts?

Comments

pfrenssen’s picture

Issue summary: View changes

Fixed HTML tag

jonathan1055’s picture

I read this post having come to the Simpletest issue queue to try to find the best place to offer an enhancement I have written. It is well written and lists the options very clearly, so I was surpised that no-one has commented on this since Aug 2012. [Then even more surprised to see it was written by my friend Pieter from Scheduler.]

I would say that (2) is the better course of action right now. We would need to start by agreeing on the contrib features which core should use, and have buy-in from the core maintainers to participate in answering questions and help with problems that arise.

Jonathan

Dave Reid’s picture

I do think that either we should shoot for merging features into core or split stuff into separate projects. The clone and remote test bases could arguably be their own separate projects that could innovate for the people interested in maintaining them.

My interest in working on this contrib was helping backport stuff to the Drupal 6 branch. As I'm no longer actively maintaining D6 contrib modules, and the amount of effort to get tests for D6 working is too great, I'm not going to really see myself contributing to this module in the future.