[dojo-contributors] 1.3 direction

Bill Keese bill at dojotoolkit.org
Fri Sep 26 22:34:50 EDT 2008

Oh, I guess by matrix testing you are talking about running the DOH 
tests (and existing robot tests) nightly (automatically) on a bunch of 

I think "the discussion" you are referring to is 
http://thread.gmane.org/gmane.comp.web.dojo.devel/9271/focus=9275 but 
note that it was about various topics, including automated and manual 

Note that the matrix/chart on 
http://developer.yahoo.com/yui/articles/gbs/ is just talking about what 
configurations are supported, not about automated tests.

Adam L. Peller wrote:
> On Fri, Sep 26, 2008 at 1:27 AM, Bill Keese <bill at dojotoolkit.org 
> <mailto:bill at dojotoolkit.org>> wrote:
>     Adam Peller wrote:
>      > Matrix testing for the tests we've got being much higher priority
>     than writing automated tests, IMHO.
>     Hmm, ISTM that given the number of OSs (Win2K, WinXP, Vista, Tiger,
>     Leopard, Linuxes) and the number of browsers (IE6, IE7, IE8, FF2, FF3,
>     Safari, Opera, Konqueror, Chrome) that it's not practical to test every
>     single combination by hand.  And it also seems really unlikely that
>     something will work correctly on IE7/vista but fail on IE7/WinXP.
> Right... that's exactly my point.  The discussion was about getting our 
> existing tests to run automated on servers in various VM (like Mozilla's 
> tinderbox).  Whether we include Windows variants is less important.  I 
> know this provides coverage for only a fraction of the features and does 
> not address the huge test issues with interactive, but it provides a 
> solid base and help us catch regressions.  I think it's the logical 
> place to start.
> ------------------------------------------------------------------------
> _______________________________________________
> dojo-contributors mailing list
> dojo-contributors at dojotoolkit.org
> http://turtle.dojotoolkit.org/mailman/listinfo/dojo-contributors

More information about the dojo-contributors mailing list