Re: [E-devel] conf_randr - A Randr dialog for e17

[Cedric B. <ced...@fr...>]

Hi,

On Tue, Nov 1, 2011 at 11:59 AM, Stefan Schmidt
<st...@da...> wrote:
> On Tue, 2011-11-01 at 19:01, Carsten Haitzler wrote:
>> On Mon, 31 Oct 2011 20:35:37 +0100 Stefan Schmidt <st...@da...>
>> said:
>>
>> > More interesting is a segfault when edje_cc builds the .edj. A
>> > backtrace points to ecore_mainloop and friends used for the SVG stuff.
>> > And indeed, converting the svg to png make edje_cc happy. (separate
>> > mail for the issue)
>>
>> separate issue indeed - but i didn't see that problem. worked for me. :/
>
> Will be tracked down separately.
>
>> > On Wed, 2011-10-19 at 21:37, Leif Middelschulte wrote:
>> > >
>> > > find attached another version of the patch. It "works" here, but I
>> > > don't have the ressources to test it properly.
>> > >
>> > > Known issues:
>> > > Frontend (conf_randr):
>> > > - not pushed into e/demodulized yet
>> > > - dialog has to be resized manually for displays to be proper positioned
>> >
>> > Thats a bit confusing at the first time. The display should be changed
>> > to something vendor neutral btw. :)
>>
>> a small issue. but there are bigger fish to fry. :)
>
> Agreed, but it asks for unneeded trouble to get it in like this.
>
>> > I only see one output here which seems to be the LVDS of my notebook.
>> > The external monitor connected over DVI and from xrandr detected as
>> > HDMI2 seems to be missing completely. Problems with EDID parsing or
>> > such? Where should I look for debug infos for this?
>>
>> did you try restart e after it's plugged in? i have it showing hdmi and lvds.
>> this i think is the issue of polling. you need to keep checking for new screens
>> with current xorgs to see a new monitor. this is something ifound last patch
>> rounf.
>
> Actually I was thinking I did restart e. Will do for sure when I test
> again in the evening. Also some more testing with VGA and also on a
> different machine with older intel graphics card.
>
>
>> > Need to find out why it does not detect my external monitor the next
>> > days.
>>
>> aaaah you haven't gotten to the next phase then... where it's a bitch to get it
>> to work at all. i've been trying to drag monitors left and right and they stick
>> then reset afterwards. if i select things like "left of" or "right of" and so
>> on i find it just doesnt switch screens on apply - dnding the mini-monitors
>> around seems pot-luck on a good day - after much clicking, selecting and
>> dragging i finally got it to configure hdmi to be left of lvds... but this
>> isn't working well right now - not well enough to "go gold". th crash - can be
>> fixed. the resizing. can be fixed. i agree with the monitor branding thing.
>> also mine are squashed and ultra-wide. but thats gfx. the real issues are in
>> this actually working well enough to iron out the lesser issues.
>
> All this has to wait until I can actually can see my different
> screens. Job for tonight. :)
>
>> leif - right now it still needs work. the ui and overall direction is right.
>> now it actuallly needs to work beyond the "roll some dice" point. well.. here
>> is also the bit where i ask you... does it work solidly for you? can u configre
>> screens to be left/right/above/below of others reliably with a quick drag +
>> apply or select from radio buttons? because it definitely doesnt work reliably
>> for me :(
>
> Not gold. That he mentioned himself. The question I ask myself is if
> it is good enough to go into svn while being disabled by default.
> Would make testing and fixing easier. No idea what Leif thinks about
> this. Not a big problem for me anyway can use the patch in a different
> branch here for all my testing.

Not to mention having more devs working on it and more people testing
it. I do expect it to be broken, but at least we can start to work
fixing the issue together. I vote again to push that in svn with
default build being disabled.
-- 
Cedric BAIL