Re: [E-devel] Ecore_X_Randr API patches

[Leif M. <lei...@gm...>]

Thanks for quick review.

2010/8/3 Carsten Haitzler <ra...@ra...>:
> On Tue, 3 Aug 2010 17:38:10 -0300 Leif Middelschulte
> <lei...@gm...> said:
>
> review (the short and curt version - forgive the brevity. trying to save time).
>
> EAPI void ecore_x_randr_external_init(Ecore_X_Display *disp);
>  why pass in the disp? all ecore_x assumes a single display connection (its
> how its designed - yes. a limitation). just make that a void for param.
I actually first put it in to test stuff outside of a module. You mean
use environment variable when _external_init() is called?

> EAPI const Eina_Bool      ecore_x_randr_query(void);
>
>  kill const.
will do.

> EINA_DEPRECATED EAPI const Ecore_X_Randr_Orientation
> ecore_x_randr_screen_primary_output_orientation_get(Ecore_X_Window root);
>
>  dont even deprecate - remove entirely. same with all other deprecated calls.
this is RandRR <= 1.1 stuff. Just marked it deprecated as usually most
drivers support >=1.2, which is a superset of 1.1 but uses different
ways to do the same stuff.

> EAPI void ecore_x_randr_screen_crop(Ecore_X_Window root);
>
>  if i am not crazy - that code doesnt corp. it UNCROPS. ie allows all screens
> to be fully seen for that root window, but keeps them cropped 1:1 resolution ->
> no? the function name at any rate doesnt really say what the function does.
> maybe some form of "ecore_x_randr_screens_reset()" is more appropriate here?
This function is used when a monitor is detached. When a CRTC is
disabled, the screen stays the size it was, which is pointless. That's
why I introduced this convenience function. And it actually just deals
1 screen's CRTCs.

> that's it for now. fix these - get patches back for ecore and e to adapt to new
> api and it can go in.

Will do within a few minutes after you comment my points above ;-)

> then add randr-multi-head stuff to e dialog :)
Looking forward to do so after this patch generation horrortrip ^^

>
>> Dear ladies and gentlemen,
>>
>> find attached a set of patches, which are splitted as much as I could
>> see reasonable. They are based on r50745. Uncrustification was done
>> with latest SVN script in FORMATTING.
>>
>> ecore:
>> - uncrustification of modified files (discomfitor's uncrustification
>> didn't work properly (back then at least))
>> - renaming/partialy altering old RandRR API <= 1.1
>> - introduction of new code for RandRR >= 1.2
>>
>> e:
>> - uncrustification of modified files
>> - adjusting current code of e and conf_display to use altered API
>>
>> The patches for e rely on the RandRR 1.1 patches for ecore being
>> already applied.
>> I'm working the last bits for nice randr integration in e, but
>> splitting up the patches was, due to me being new to SVN<->Git and
>> uncrustify, a little fight and took way more time then I intended to
>> spend on it.
>> I hope this gets reviewed quick, so I can finaly use SVN head again
>> without merging horror because of formatting stuff.
>>
>> BR,
>>
>> Leif Middelschulte
>
>
> --
> ------------- Codito, ergo sum - "I code, therefore I am" --------------
> The Rasterman (Carsten Haitzler)    ra...@ra...
>
>