Status of the "Compose in a tab" experiment

Jonathan Protzenko jonathan.protzenko at gmail.com
Tue Apr 5 21:44:42 UTC 2011


Well the discussion is not uninteresting, and we're on tb-planning, 
which is supposed to be the right kind of place to discuss this :-). I 
probably under-estimated the importance of XUL. You're making good 
points, especially concerning <command> and <key>, which indeed have no 
good equivalent in JS.

On Tue 05 Apr 2011 10:34:49 PM CEST, Jim wrote:
> On Tue, Apr 5, 2011 at 3:11 PM, Jonathan Protzenko
> <jonathan.protzenko at gmail.com>  wrote:
>> My point is many people are more proficient with HTML than with XUL. Andy
>> Chung, for instance, does nothing but HTML designs. And he's using
>> -moz-box-flex, which *is* the XUL box model. What's left that XUL has and
>> HTML hasn't?<xul:description crop="end">  to emulate text-overflow: ellipsis
>> that gecko *still* hasn't implemented.  What else? Menus, but these will
>> come somehow with HTML5. A few native widgets...
> 
> The native widgets are pretty important for me, since things like
> splitters, group boxes, notification boxes, progress meters, etc. are
> all pretty useful in Thunderbird. Additionally, XUL<key>s are a
> pretty simple way to implement keyboard accelerators. The last time I
> tried to do something like that in HTML, I just gave up. Relatedly,
> the<command>  tag seems pretty useful for intelligently
> enabling/disabling multiple UI elements (menu items, buttons) that act
> on the same command.<preference>  tags seem pretty useful, though I
> haven't used them much, and you could probably implement those fairly
> easily with some Javascript anyway.
> 
> There are certainly more useful UI elements in HTML5 (though not all
> of them have been implemented in Gecko), and some of the ones I
> mentioned above are probably due to be added (I believe progress
> meters are in HTML5, but not implemented yet), but there are still
> some significant gaps which I'm not sure will ever be closed by HTML.
> 
>> There's little fresh blood coming to the Thunderbird world. I don't think we
>> can afford to be picky. Anyone who's talented with HTML could be help on
>> this project, imho :-).
> 
> Well, from my perspective, learning to work in XUL (while having HTML
> experience) was one of the easier things about contributing to
> Thunderbird. I think in terms of getting more people to contribute,
> there are bigger issues than XUL. High up on the list for me are
> writing tests (I still just copy-and-paste from other tests) and
> getting Thunderbird builds set up, which is non-trivial especially if
> you're not on Linux.
> 
> I hesitate to post too much more on this, since it's veering
> dangerously off the original topic, which was about how
> compose-in-a-tab was coming along. :)
> 
> - Jim



More information about the tb-planning mailing list