I wasn't able to reproduce this either, but I committed a patch which gets rid of the artifacts in ImageSelection-test.
Eric On 2011-09-06, at 6:39 PM, Germán Arias wrote: > I will check the gorm files. Anyway I'm thinking reimplement the panel. > If I will add more elements in each release, will be so difficult > maintain the gorm file. Thanks. > > On mar, 2011-09-06 at 22:47 +0200, Fred Kiefer wrote: >> I think Eric is correct here. I checked the Grom file and the buttons >> are positioned on half pixels. We need to change the code to get more >> reliable results here, but for now it seems easier to adjust FisicaLab. >> >> Fred >> >> On 06.09.2011 20:44, Eric Wasylishen wrote: >>> I haven't had a look at Fisicalab yet, but something which may or may not >>> be related is the artifacting I get in GSTest/ImageSelection-test. It >>> happens when moving the sliders to adjust the image size; the image can >>> leave lines as it's being resized, but the same code works fine on OS X. >>> >>> The artifacting seems to to come from calling [aView >>> setNeedsDisplayInRect:] with a non-integral rect. I suspect at some point >>> the rect coordinates are being floor()'ed, which would be incorrect. If we >>> mark a (possibly non-integral) region of a view as needing to be redrawn, >>> at some point we'll need to compute a pixel aligned version that completely >>> encloses the original rect. >>> >>> -Eric >>> >>> On 2011-09-05, at 3:47 AM, Fred Kiefer wrote: >>> >>>> On 05.09.2011 02:26, Germán Arias wrote: >>>>> Currently when I select an element in FisicaLab an ugly lines appears. >>>>> See attached image. I'm using cairo backend. >>>> >>>> I tried to reproduce this behaviour with Fisicalab 0.2.1 and current SVN >>>> GNUstep code and failed to get these artefacts. I am using the cairo >>>> backend as well. >>>> >>>> Could you please give some more details on how to reproduce this? >> >> >> _______________________________________________ >> Gnustep-dev mailing list >> [email protected] >> https://lists.gnu.org/mailman/listinfo/gnustep-dev > > > _______________________________________________ > Gnustep-dev mailing list > [email protected] > https://lists.gnu.org/mailman/listinfo/gnustep-dev _______________________________________________ Gnustep-dev mailing list [email protected] https://lists.gnu.org/mailman/listinfo/gnustep-dev
