desktop e-mail UI using gaia e-mail libs (was: TB and AMO...)

Andrew Sutherland asutherland at asutherland.org
Mon Oct 28 03:16:31 UTC 2013


On 10/27/2013 09:59 AM, neandr at gmx.de wrote:
> Very clear if MoCo is moving more and more to JS/HTML/CSS based 
> technique the XUL will become obsolete. From the perspective of a 
> contributor for legacy add-ons (written with XUL) this is an important 
> point for future offerings.

I wouldn't call it a MoCo thing, more a Firefox/Gecko thing and a marked 
increase in the capabilities of HTML for UI layout purposes (both 
functionally and from a performance perspective).  Neither XUL nor XBL 
will ever be a web standard.  It makes sense to build Firefox's UI using 
web standards technologies to the maximum extent possible, both for 
philosophical and pragmatic reasons (less code, better performance).

The web standards version of XBL which is half of what powers XUL is the 
Web Components effort (http://www.w3.org/TR/components-intro/). The 
other half of XUL is native C++ code, some of which is layout stuff 
related to moz-box that is obsoleted by CSS3 flexbox 
(https://developer.mozilla.org/en-US/docs/Web/Guide/CSS/Flexible_boxes). 
Web Components in conjunction with the flexbox stuff can be made to do 
everything XUL does.

The main question is whether anyone would undertake a direct 
porting/re-implementation of XUL or whether they would create something 
new that takes what Firefox actually needs/wants from XUL.  My guess 
would be the Firefox team would not pursue porting XUL, especially since 
HTML5 has subsumed some of the essentials XUL provides (ex: HTML5 
menu/menuitem 
http://www.whatwg.org/specs/web-apps/current-work/multipage/interactive-elements.html#the-menu-element), 
but it would not surprise me if there was a community effort to do so.  
Alternatively, a lot of XUL could probably be mapped onto existing HTML5 
elements or new WebComponents bindings using XSLT, etc.


> Any ideas about plans of the XUL lifetime?

I think XUL will last for a looong time, but like you, I think it will 
be used less and less within Firefox proper.  Since XUL can no longer be 
used for internet pages unless manually whitelisted, it's then really a 
question of when the maintenance burden of the XBL and XUL C++ code 
becomes too much of a hassle and it is removed from Gecko.  I think it 
can easily live on via web components if someone steps up to do that.  
(And I think that's much more tractable than trying to maintain the C++ 
code in comm-central, etc.)

In general, I wouldn't worry about it.  Web Components are still under 
heavy development and there is XUL even in Firefox OS.  It will be quite 
some time before anyone can begin to propose removing XUL with any 
seriousness.  (And when they do, it will probably turn out there are 
another 10 fundamental things that depend on it!)

Andrew




More information about the tb-planning mailing list