W3C home > Mailing lists > Public > public-css-testsuite@w3.org > April 2012

Re: Submission of transform tests

From: Linss, Peter <peter.linss@hp.com>
Date: Sun, 15 Apr 2012 20:42:36 +0100
To: Aryeh Gregor <ayg@aryeh.name>
CC: "public-css-testsuite@w3.org" <public-css-testsuite@w3.org>
Message-ID: <F97C89C4-5283-4EAA-AB17-101222F53BAD@hp.com>

On Apr 14, 2012, at 11:44 PM, Aryeh Gregor wrote:

> On Sun, Apr 15, 2012 at 9:18 AM, Aryeh Gregor <ayg@aryeh.name> wrote:
>> Should I prefix the base name with "transform-" or such?  If the test
>> name is expected to be globally unique, I don't want to steal
>> globally-relevant names for very specific tests.
> 
> And indeed, Shepherd got very confused, because there were naming
> conflicts.  See <http://test.csswg.org/shepherd/testcase/abspos-001>
> et al.  I've renamed them to start with "transform-".  Now I hope
> they're ready for review again.

Actually Shepherd worked as designed there. Shepherd tracks each asset by name, so if a file's name changes it's considered a new asset. It does understand that the same asset may be represented by multiple files (as commonly happens with support files, and sometimes with references and tests as they're being worked on) and it tracks each file independently, but they're all considered the same asset. 

I do have some improvements on how Shepherd handles file renames on my to do list (it's all still a work in progress). When comments are entered they are logged against a specific revision of a specific file, so I can have the comments follow the file in the case of renames.
Received on Sunday, 15 April 2012 19:47:37 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Sunday, 15 April 2012 19:47:45 GMT