On Sat, Mar 7, 2009 at 3:26 AM, Sami Jaber <[email protected]> wrote:

> I think it could really (really) be interesting to get OOPHM with 1.6, is
> it so much work John (why not using a generator for the overlay classes,
> Sanjiv seems to use this kind of tool for SmartGWT) ?
>
> I don't know about SmartGWT, but basically what you are asking for is
automated merge conflict resolution, and that just isn't feasible.  For
example, when Bob added support for JSNI class listerals with ::class, Joel
and I had to dig around in the legacy hosted implementation and the OOPHM
overlay classes to figure out what needed to be done, and it wasn't solved
the same way.

> I was using OOPHM as a early adopter since the beginning and with the
> incompatibilities between OOPHM and the new war directory layout, there is
> no way to provide you for feedback and I have lost in term of productivity
> (compile time is still so slow in 1.6)
>

As we get closer to the next release, we will get OOPHM running with
HostedMode (and therefore war) and make it the sole way of running hosted
mode.  However, there are a number of UI improvements that need to be made
before we can do away with legacy hosted mode (plus some other issues like
the testing story on Windows/Mac - in Linux we can run a Selenium-RC server
inside Xvfb, but there is no equivalent on Windows/Mac), and right now there
are competing priorities.  Adding additional work to try and make it usable
with 1.6 seems counterproductive to actually getting it to its releasable
form.

> btw, runAsync was slated for the release after 1.6 and it is still in the
> trunk for a while now, does it mean that the scope is being changed ?
>

The trunk *is* for the release after 1.6, so the big things for that release
will be runAsync, SOYC, and OOPHM.

-- 
John A. Tamplin
Software Engineer (GWT), Google

--~--~---------~--~----~------------~-------~--~----~
http://groups.google.com/group/Google-Web-Toolkit-Contributors
-~----------~----~----~----~------~----~------~--~---

Reply via email to