[Qtscript-bindings] long qtscript binding load times: easily broken up?

Peter Zhou peterzhoulei at gmail.com
Sat Sep 13 03:45:41 CEST 2008


This is awesome!

On Sat, Sep 13, 2008 at 12:03 AM, Kent Hansen <khansen at trolltech.com> wrote:
> Richard Moore wrote:
>> On 9/10/08, Ian Monroe <ian at monroe.nu> wrote:
>>
>>> I was thinking of maybe a combination of breaking up the modules into
>>> smaller ones and disabling classes that don't sound useful for Amarok
>>> scripts (and then re-enable them on-request).
>>>
>>> Other solutions? How much should I break up? I could potentially have
>>> a "each class has its own module." Would this incur a lot of overhead
>>> loading all these modules or does it not make much difference?
>>>
>>
>> Kent and I discussed this at Akademy, I believe he's already got some
>> code for a version of the generator that creates bindings that do on
>> demand loading to remove the overhead.
>>
>> Cheers
>>
>> Rich.
>>
>
> Yes, I've got a patch brewing. You can track the thing here:
> http://code.google.com/p/qtscriptgenerator/issues/detail?id=10
> The patch still keeps the classes in libs that match Qt's libraries
> (e.g. all of bindings for QtGui in one library), but it cuts down
> drastically on initialization time since it only creates "stubs" for all
> the binding classes.
>
> Kent
> _______________________________________________
> Qtscript-bindings mailing list
> Qtscript-bindings at kde.org
> https://mail.kde.org/mailman/listinfo/qtscript-bindings
>



-- 
Peter
-------------------------------
http://www.peterzl.net/


More information about the Qtscript-bindings mailing list