W3C home > Mailing lists > Public > public-webapps@w3.org > January to March 2013

Re: Proposal: moving tests to GitHub

From: Rebecca Hauck <rhauck@adobe.com>
Date: Thu, 31 Jan 2013 12:18:57 -0800
To: Odin HÝrthe Omdal <odinho@opera.com>, Arthur Barstow <art.barstow@nokia.com>
CC: "public-webapps@w3.org" <public-webapps@w3.org>
Message-ID: <CD30064E.26816%rhauck@adobe.com>
I guess I should chime in.

Yes I submitted a batch of tests from TestTWF some time after the Paris
event. After having a pretty bad experience with Mercurial earlier in the
year at TestTWF San Francisco, we made a conscious choice to eliminate it
in Paris and use DropBox instead. It was marginally better if only for the
fact that people were actually writing tests rather than futzing with
Mercurial for half the day (this happened at the SF event - it got very
ugly for some).

The main flaw in the DropBox process was getting the tests moved to the
respective repositories. It was work we recognized and agreed to do up
front in favor of getting more people writing tests.  We first reached out
to all the test authors individually, gave them the Mercurial info, and
offered assistance getting the tests submitted (hence the time lag). For
whatever reasons, very few actually took the steps to submit themselves.
Some were even under the impression that the tests would be submitted for
them, a reasonable assumption based on the W3C Grant of License they
signed at the event.  So, to not lose those tests, I ended up submitting
them on behalf of the authors.

Since I'm not in the webapps working group, I had to first get access to
the repository. I was told that that to get write access, I (probably) had
to join the working group [1]. So it's sort of by definition that there
are no "outsider" submissions in dvcs, no?  Either way, after several
emails with several people, after about 5 days, I got write access and
pushed. I'm a semi-insider and even my experience wasn't that great. And
yes Odin, I completely agree that direct submission through pull requests
at the event would definitely have been more engaging with newcomers. I'm
pretty sure this will be addressed correctly at a future event.

And for lack of a better tracking mechanism, I added a <!-- Submitted from
TestTWF Paris --> comment to each tests and in some cases, put them in a
similarly named folder in the repo. If you grep the webapps test dir,
you'll find more than one. :)

-Rebecca

[1] 
http://lists.w3.org/Archives/Public/public-test-infra/2012OctDec/0025.html

On 1/31/13 9:47 AM, "Odin HÝrthe Omdal" <odinho@opera.com> wrote:

>On Thu, 31 Jan 2013 18:13:15 +0100, Arthur Barstow
><art.barstow@nokia.com>
>wrote:
>> However, AFAIK, currently, only one of WebApps' thirty active specs
>> actually has an "outside" contribution.
>
>I should've already left work, so I'll just reply to this sentence
>quickly  
>:-)
>
>With that you mean Server Sent Events?  The /only/ webapps spec (that I
>know of) that actually is at GitHub and not at w3c dvcs?  In that case we
> 
>have "outside" contributions on 100% of the tests on GitHub, and 0% on
>the  
>tests at w3c dvcs.
>
>Although, I don't really know how to count TestTWF, but there's quite a
>lot of webapps tests there as well.  Rebecca Hauck pushed them to dvcs a
>while ago, although that was a good time after the event.  I'll hazard to
> 
>guess that if they could do the pull requests right there at the event
>(like what happened for Server Sent Events), we might have a higher
>retain  
>rate.  Because I think it'd make people feel easier/more involved.
>Speculation, but not totally unfounded :-)
>
>
>Many other good points. I guess Robin, James, Tobie or someone else will
>already have replied by tomorrow. :]
>-- 
>Odin HÝrthe Omdal (Velmont/odinho) ∑ Core, Opera Software,
>http://opera.com
>
Received on Thursday, 31 January 2013 20:16:39 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 18:49:57 GMT