On Fri, Jan 29, 2016 at 5:02 AM, Cedric BAIL <[email protected]> wrote: > On Fri, Jan 29, 2016 at 4:00 AM, Jean-Philippe André <[email protected]> > wrote: >> On 29 January 2016 at 08:27, Cedric BAIL <[email protected]> wrote: >>> On Thu, Jan 28, 2016 at 3:12 PM, Stefan Schmidt <[email protected]> >>> wrote: >>> > On 10/12/15 21:56, Mike Blumenkrantz wrote: >>> >> discomfitor pushed a commit to branch master. >>> >> >>> http://git.enlightenment.org/core/elementary.git/commit/?id=6149fd7a0a9ef3f9495270cf1c4ab7727a30f21c >>> >> >>> >> commit 6149fd7a0a9ef3f9495270cf1c4ab7727a30f21c >>> >> Author: Mike Blumenkrantz <[email protected]> >>> >> Date: Thu Dec 10 15:53:31 2015 -0500 >>> >> >>> >> win: redo resize deferral to trigger during pre render >>> >> >>> >> using a job results in the resize occurring after render in the >>> same loop, >>> >> meaning that anything which depends on resizes to happen in a >>> certain >>> >> time will not have synchronized display vs toolkit geometries >>> >> >>> >> fixes all issues related to wayland window geometries >>> > >>> > This commit breaks the widget screenshot generation for our Elementary >>> > docs. Two example from our nightly doc generation: >>> > >>> > >>> https://build.enlightenment.org/job/nightly_elm_gcc_x86_64/lastSuccessfulBuild/artifact/doc/html/group__Button.html >>> > >>> https://build.enlightenment.org/job/nightly_elm_gcc_x86_64/lastSuccessfulBuild/artifact/doc/html/group__Flip.html >>> > >>> > As you can see the image are off or completely broken. They should >>> > actually look like this (from our 1.16 docs) >>> > >>> > https://docs.enlightenment.org/elementary/current/group__Button.html >>> > https://docs.enlightenment.org/elementary/current/group__Flip.html >>> > >>> > Thanks to Jean Rene Dawin for reporting it. It took me a long time to >>> > git bisect'ing it down to this commit. Reverting it locally fixes the >>> > problem. >>> > >>> > I'm open for suggestions how we fix this and still fix the problem the >>> > commit originally tried to fix. >>> >>> Relying on render pre make sense in this patch. I am wondering how >>> exactness and our doc subsystem do the screenshot. Do you have a >>> pointer to it ? >>> >>> >> make doc for elementary relies on the "shot" engine. I don't know the exact >> wrt. this patch, though. > > There is something common in that case, both exactness and the shot > engine use ecore_evas_manual_render. I can't see why there would be > something wrong in the case of a manual render. It seems to call the > render_pre function correctly and the buffer backend is not async, so > it should be filling the buffer correctly. Also the image are of the > right size, which means that the width and height have correctly been > set on the ecore_evas before we do the pixels get. Going to try > playing with it and see if I find something.
Ok, this seems to have been way easier than I though. The pre render function wasn't called in the buffer engine (This code should really be refactorised one day...). I fixed it with commit adff529004b2d8b4c2395c74af7f8f86a2a42ebc. Please test and let me know if you still see issues. -- Cedric BAIL ------------------------------------------------------------------------------ Site24x7 APM Insight: Get Deep Visibility into Application Performance APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month Monitor end-to-end web transactions and take corrective actions now Troubleshoot faster and improve end-user experience. Signup Now! http://pubads.g.doubleclick.net/gampad/clk?id=267308311&iu=/4140 _______________________________________________ enlightenment-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
