[Webkit-unassigned] [Bug 27428] adding "base class" GdomDOMObject for GObject bindings

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=27428





--- Comment #11 from Luke Kenneth Casson Leighton <lkcl at lkcl.net>  2009-08-08 02:32:42 PDT ---
(In reply to comment #10)
> This is not using the WebKit prefix for the DOM bindings that was agreed in bug
> #16401, may I know why? 

 yes of course. 

 i've explained a number of times, but am happy to explain again.

 1) it's what alp originally chose - i just went along with it when i started
this work in august 2008.

 2) the work completed to a satisfactory milestone, i left it alone.
    much discussion ensued, requesting backtracking and removal of
    critical features that were part of meeting the original milestone.

    due to the implications of these request [mostly that full testing
    would not be possible] i declined.

 3) adam took over, and complied with the request to remove critical
    features.

 4) this _immediately_ excluded the test platform which had been utilised
    to perform extensive empirical validation of the work.

    thus, i was excluded from contributing any code, but i carried on
    participating in providing insights and advice.

 5) a (sensible) decision was made to rename Gdom to Webkit, GDOM to WEBKIT
    etc. as part of those discussions (as you've observed, and i was
    part of, and i agree with)

 6) adam's version of the patch was completed to a satisfactory milestone
    for yorba's needs.

    the discrepancy between this and the milestone reached in 2) was... vast.

 6) yorba instructed adam to cease further work.

 7) i reached an agreement with david that allowed further work to continue
    using sensible tools (multiple bugreports).

 8) the gap between the code that i understood - and, much more importantly,
    could _validate_, was so large that i had to make a decision to return
    to the _original_ patch.

 9) i posted a message expanding on 8) to the #16401 bugtracker, and awaited
    responses.

 10) no responses were received.

 11) perhaps more importantly, no offers of help were received either.

 12) i therefore took the simple pragmatic approach, which was that if
     people weren't going to pay attention, or offer to help, it was
     probably best just get on with it.

 13) once all the (15, due to be 30) patches are landed, i figured that
     then would be a good time to do the renaming, as it would then be
     tracked, managed, discussed; there would be svn diffs, revision
     history etc. etc.

 so, the summary is: due to wholly practical considerations, and due to lack
 of community input, i made a decision to get on with the job.


> It also was committed using ChangeLog entries from a
> completely different bug.

 acchh. darnit.  i'm maintaining 15 separate patches (in the same svn) at the
moment - that's due to go up to 30, today.  i'll be writing a script which
helps manage this, today, so that ChangeLog entries go at the top of the file
etc.

 sorry.

l.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.