See also: IRC log
"Send email to the public-html-testsuite mailing list describing the test cases you are submitting as well as giving permission for their use as described in the licensing section above"
http://www.w3.org/2002/09/wbs/1/testgrants2-200409/
plh: folks need to go through the grant of license for legal reasons
kris: we'll watch the mailing list and tell people to fill the form if necessary
Send email to the public-html-testsuite mailing list describing the test cases you are submitting as well as giving permission for their use as described in the licensing section above, using the grant of license.
(that's the new sentence)
plh: we're covered for participants of the working group already. no need to fill the form.
kris: so I can commit tests directly into mercurial?
plh: yes
kris: what about licensing information?
<paulc> Can we reword the following: " For example you should treat all browsers the same, no special logic, test dependencies should be supported by all browsers."
<paulc> For example you should treat all browsers the same. Your tests should not use any special logic for a specific browser."
<scribe> ACTION: plh to follow up on licensing information in test suite files [recorded in http://www.w3.org/2010/03/09-htmlt-minutes.html#action01]
<paulc> From http://www.w3.org/html/wg/wiki/Testing/Submission/
Paul: [proposed wording for last sentence]
Kris: accepted
http://www.w3.org/html/wg/wiki/Testing/Feedback/
plh: we need to give directions on the kind of
information needed in order to get proper feedback
... such as which test and which version of the test
kris: seems fine to me
paul: why don't they file a bugzilla report on the test suite instead?
kris: would be fine
paul: it does require an account, but thinks
about the wording in the working drafts.
... the more people that do this by filling bugs, the easier it will be.
... we should treat the test suite as a product of the working group
... and if we discover that there is an error in the spec, we can file a
bugzilla on the spec that points to the bugzilla on the test
<krisk> OK - I'll update the wiki with this information
kris: not sure why bugzilla didn't use my w3c credentials
paul: w3c isn't using a federated identity system
plh: correct, our system folks are implementing one (mercurial is an example of the future system)
paul: there are in fact two bugs for the test
suite in there
... reuse text from the working draft status section
http://www.w3.org/html/wg/wiki/Testing/Approval/
Paul: question: "Anyone in the HTML working group
with the ability to read and understand the spec and a thorough understanding
of how to write a good test case can review additions and changes to the
Approved collection."
... how do they declare their position?
... is there some sort of voting mechanism?
... how to we record the process?
Kris: I would send a CfC on a set of test cases?
<paulc> Is there a min bar ie. at least one reviewer?
Paul: so no minimum bar to have at least one person to review the tests?
Kris: I'd like to work that way: someone sends a test, get some feedback, make some changes, and then request for approval. the third party would say "looks good"
paul: the page doesn't say that
... btw, is it the task force that approves the tests or does the TF need to
go back to the WG?
... best is to have a proposed mechanism, and ask permission from the WG.
after that, you'll be ok.
... but you have to anticipate that somebody in the WG is going to be
reluctant to engage in the TF and start a thread on the WG list instead
... the chairs can then assign the matter to the TF
... polish the 3 pages, then at some point, ask approval from the WG about the
procedures established in the TF
... but the text needs to be more explicit
kris: agreed. needs more of the how, not just the who