I've pushed some patches of Yura's to infusion and to our old node-jqUnit repository which brings them into
sync and regularises a few issues relating to relative paths, as well as updating to the latest trunk of
QUnit. I could verify that they were basically working by manually updating a few of GPII's test cases -
however, to enable GPII to move up to these versions of the modules, we will need to make progress on our
support for "require" in the browser which was temporarily removed as a result of the discussion on this JIRA:
http://issues.fluidproject.org/browse/FLUID-4911
I've updated the JIRA with some commentary and references reflecting the discussions we had in San Diego. As
far as I can see, our old implementation named "requireStub" needs to be revived, and updated with the use
of a "synchronous AJAX" approach inspired by "Jay's" very short require implementation. In the long term, we
should consider a move to AMD/require.js once we have managed to make progress on Infusion's and GPII's
builder most likely using "grunt". We don't care about the performance impact of synchronous AJAX on the
client since this implementation is only for testing purposes - and we don't care to primarily write test
cases in node.js where it is not necessary because the node-inspector project has been drifting without a
maintainer for nearly a year. However we do wish to ensure that all GPII test cases written in the browser
will also run in node.js.
Please respond if you do or don't consider this is a reasonable summary of the point our reasoning had taken
us - or with any other concerns or questions.
Cheers,
Antranig
_______________________________________________________
fluid-work mailing list - [email protected]
To unsubscribe, change settings or access archives,
see http://lists.idrc.ocad.ca/mailman/listinfo/fluid-work
