October 2015 - Pharo-dev - lists.pharo.org

Re: [Pharo-dev] Yet another Athens font thread

by Aliaksei Syrel

Also "broken" and working scaled fonts are the same in inspector Cheers, Alex On Thu, Oct 8, 2015 at 6:05 PM, Aliaksei Syrel <alex.syrel(a)gmail.com> wrote: > Do you remember a hack with scaling? (scaleBy: 1.0001(a)1.0001) It makes > letters to have the same height, so no more huge ones with small fonts. But > it doesn't fix a bug when letters have wrong width so there is a space > between characters. The same effect can be achieved if we set font size > manually after setting scaled font before drawing glyphs with cairo. It > means that at some point we get wrong calculated glyphs. So if we "fix" > scaled font before glyph calculation - the problem would be gone. In the > patch I set scaled font, manually apply font size, retrieve "fixed" scaled > font and calculate glyphs based on it. > > Also after investigation I found something interesting. If we clean real > font of a logical font the problem gets fixed for some period of time. I > assume that FreeTypeFont gets broken after a while and logical font always > returns broken real font. Setting realFont to nil forces logical font to > create a new working free type font. As soon as new free type font is > created cairo produces scaled font and caches it. So actual problem lies > deep in FreeTypeFont... > > I even opened two images: one with broken fonts and one with working. Then > I compared in inspector two free type fonts - they are absolutely the same > and have identical instance variable in a whole object tree, except of > handler pointers to C structure which I can't inspect from pharo... > > Saving image fixes fonts because of > >> LogicalFont>>shutDown: quitting >> self allSubInstances do: [:i | i clearRealFont]. > > > Cheers, > Alex > > On Thu, Oct 8, 2015 at 5:43 PM, Nicolai Hess <nicolaihess(a)web.de> wrote: > >> How did you fix it? >> Am 08.10.2015 17:06 schrieb "Aliaksei Syrel" <alex.syrel(a)gmail.com>: >> >>> Hi >>> >>> It looks like I fixed that weird text rendering problem... Just get the >>> latest Athens-Cairo. >>> >>> Cheers, >>> Alex >>> >>> On Mon, Aug 3, 2015 at 9:32 PM, Nicolai Hess <nicolaihess(a)web.de> wrote: >>> >>>> >>>> >>>> 2015-06-10 12:01 GMT+02:00 Aliaksei Syrel <alex.syrel(a)gmail.com>: >>>> >>>>> Hi >>>>> >>>>> I noticed that if I force use "Source Code Pro" as standard font for >>>>> all elements in the image I never have problems with font rendering. In >>>>> case of "Source Sans Pro" font is just not usable becasue of rendering bugs >>>>> on mac, windows and linux(?). If "Lucida Sans Pro" is used - no bugs too. >>>>> >>>>> Because bug is reproducable on all platform - it's platform >>>>> independent bug. >>>>> Because it's reproducable only with one specific font - maybe it's not >>>>> an athens issue but bug in font itself and if someone could change/update >>>>> Source Sans Pro to a newly version or to a similar font the problem will be >>>>> solved? >>>>> >>>> >>>> Did you try to use a new font version ? >>>> I don't know exactly how embedded fonts are included. I tried to create >>>> the font with a newer version but it did not work ( or at least I could >>>> not see any difference). >>>> >>>> >>>> >>>>> >>>>> Cheers, >>>>> Alex >>>>> >>>> >>>> >>> >

Oct. 8, 2015

[pharo-project/pharo-core]

by GitHub

Branch: refs/tags/50375 Home: https://github.com/pharo-project/pharo-core

Oct. 8, 2015

[pharo-project/pharo-core] cc9766: 50375

by GitHub

Branch: refs/heads/5.0 Home: https://github.com/pharo-project/pharo-core Commit: cc976699b65f50807b3015a8a31b1663281182b3 https://github.com/pharo-project/pharo-core/commit/cc976699b65f50807b3015a8… Author: Jenkins Build Server <board(a)pharo-project.org> Date: 2015-10-08 (Thu, 08 Oct 2015) Changed paths: R Kernel.package/Object.class/instance/user interface/addModelMenuItemsTo_forMorph_hand_.st R Morphic-Base.package/MorphicModel.class/instance/menu/addCustomMenuItems_hand_.st M Morphic-Widgets-FastTable.package/FTSelectionStrategy.class/instance/private/toggleRowIndex_.st A Reflectivity.package/ReflectiveMethod.class/instance/forwarding/doesNotUnderstand_.st A Reflectivity.package/ReflectiveMethod.class/instance/forwarding/flushCache.st A Reflectivity.package/ReflectiveMethod.class/instance/forwarding/methodClass_.st A Reflectivity.package/ReflectiveMethod.class/instance/forwarding/selector.st A Reflectivity.package/ReflectiveMethod.class/instance/forwarding/selector_.st R Reflectivity.package/ReflectiveMethod.class/instance/reflective operations/doesNotUnderstand_.st R ScriptLoader50.package/ScriptLoader.class/instance/pharo - scripts/script50374.st A ScriptLoader50.package/ScriptLoader.class/instance/pharo - scripts/script50375.st R ScriptLoader50.package/ScriptLoader.class/instance/pharo - updates/update50374.st A ScriptLoader50.package/ScriptLoader.class/instance/pharo - updates/update50375.st M ScriptLoader50.package/ScriptLoader.class/instance/public/commentForCurrentUpdate.st Log Message: ----------- 50375 16542 FTSelectionStrategy>>toggleRowIndex: logic duplication https://pharo.fogbugz.com/f/cases/16542 16744 ReflectiveMethod: make sure to not use DNU forwarding on invalidate https://pharo.fogbugz.com/f/cases/16744 16741 remove addModelMenuItemsTo:forMorph:hand: https://pharo.fogbugz.com/f/cases/16741 http://files.pharo.org/image/50/50375.zip

Oct. 8, 2015

Re: [Pharo-dev] Yet another Athens font thread

by Aliaksei Syrel

Do you remember a hack with scaling? (scaleBy: 1.0001(a)1.0001) It makes letters to have the same height, so no more huge ones with small fonts. But it doesn't fix a bug when letters have wrong width so there is a space between characters. The same effect can be achieved if we set font size manually after setting scaled font before drawing glyphs with cairo. It means that at some point we get wrong calculated glyphs. So if we "fix" scaled font before glyph calculation - the problem would be gone. In the patch I set scaled font, manually apply font size, retrieve "fixed" scaled font and calculate glyphs based on it. Also after investigation I found something interesting. If we clean real font of a logical font the problem gets fixed for some period of time. I assume that FreeTypeFont gets broken after a while and logical font always returns broken real font. Setting realFont to nil forces logical font to create a new working free type font. As soon as new free type font is created cairo produces scaled font and caches it. So actual problem lies deep in FreeTypeFont... I even opened two images: one with broken fonts and one with working. Then I compared in inspector two free type fonts - they are absolutely the same and have identical instance variable in a whole object tree, except of handler pointers to C structure which I can't inspect from pharo... Saving image fixes fonts because of > LogicalFont>>shutDown: quitting > self allSubInstances do: [:i | i clearRealFont]. Cheers, Alex On Thu, Oct 8, 2015 at 5:43 PM, Nicolai Hess <nicolaihess(a)web.de> wrote: > How did you fix it? > Am 08.10.2015 17:06 schrieb "Aliaksei Syrel" <alex.syrel(a)gmail.com>: > >> Hi >> >> It looks like I fixed that weird text rendering problem... Just get the >> latest Athens-Cairo. >> >> Cheers, >> Alex >> >> On Mon, Aug 3, 2015 at 9:32 PM, Nicolai Hess <nicolaihess(a)web.de> wrote: >> >>> >>> >>> 2015-06-10 12:01 GMT+02:00 Aliaksei Syrel <alex.syrel(a)gmail.com>: >>> >>>> Hi >>>> >>>> I noticed that if I force use "Source Code Pro" as standard font for >>>> all elements in the image I never have problems with font rendering. In >>>> case of "Source Sans Pro" font is just not usable becasue of rendering bugs >>>> on mac, windows and linux(?). If "Lucida Sans Pro" is used - no bugs too. >>>> >>>> Because bug is reproducable on all platform - it's platform independent >>>> bug. >>>> Because it's reproducable only with one specific font - maybe it's not >>>> an athens issue but bug in font itself and if someone could change/update >>>> Source Sans Pro to a newly version or to a similar font the problem will be >>>> solved? >>>> >>> >>> Did you try to use a new font version ? >>> I don't know exactly how embedded fonts are included. I tried to create >>> the font with a newer version but it did not work ( or at least I could >>> not see any difference). >>> >>> >>> >>>> >>>> Cheers, >>>> Alex >>>> >>> >>> >>

Oct. 8, 2015

Re: [Pharo-dev] Yet another Athens font thread

by Nicolai Hess

How did you fix it? Am 08.10.2015 17:06 schrieb "Aliaksei Syrel" <alex.syrel(a)gmail.com>: > Hi > > It looks like I fixed that weird text rendering problem... Just get the > latest Athens-Cairo. > > Cheers, > Alex > > On Mon, Aug 3, 2015 at 9:32 PM, Nicolai Hess <nicolaihess(a)web.de> wrote: > >> >> >> 2015-06-10 12:01 GMT+02:00 Aliaksei Syrel <alex.syrel(a)gmail.com>: >> >>> Hi >>> >>> I noticed that if I force use "Source Code Pro" as standard font for all >>> elements in the image I never have problems with font rendering. In case of >>> "Source Sans Pro" font is just not usable becasue of rendering bugs on mac, >>> windows and linux(?). If "Lucida Sans Pro" is used - no bugs too. >>> >>> Because bug is reproducable on all platform - it's platform independent >>> bug. >>> Because it's reproducable only with one specific font - maybe it's not >>> an athens issue but bug in font itself and if someone could change/update >>> Source Sans Pro to a newly version or to a similar font the problem will be >>> solved? >>> >> >> Did you try to use a new font version ? >> I don't know exactly how embedded fonts are included. I tried to create >> the font with a newer version but it did not work ( or at least I could >> not see any difference). >> >> >> >>> >>> Cheers, >>> Alex >>> >> >> >

Oct. 8, 2015

Re: [Pharo-dev] [OT] Slides the web way

by Esteban A. Maringolo

2015-10-08 8:42 GMT-03:00 Damien Cassou <damien.cassou(a)inria.fr>: > > Esteban A. Maringolo <emaringolo(a)gmail.com> writes: > >> I used reveal.js vÃa Slides.com for a demo [1] >> >> It's produces really nice presentations. > > could you please give us the sources? I think I can't... because I use the free plan. :-/ I edited everything using the slides.com GUI editor. Regards! Esteban A. Maringolo

Oct. 8, 2015

Re: [Pharo-dev] Yet another Athens font thread

by Aliaksei Syrel

Hi It looks like I fixed that weird text rendering problem... Just get the latest Athens-Cairo. Cheers, Alex On Mon, Aug 3, 2015 at 9:32 PM, Nicolai Hess <nicolaihess(a)web.de> wrote: > > > 2015-06-10 12:01 GMT+02:00 Aliaksei Syrel <alex.syrel(a)gmail.com>: > >> Hi >> >> I noticed that if I force use "Source Code Pro" as standard font for all >> elements in the image I never have problems with font rendering. In case of >> "Source Sans Pro" font is just not usable becasue of rendering bugs on mac, >> windows and linux(?). If "Lucida Sans Pro" is used - no bugs too. >> >> Because bug is reproducable on all platform - it's platform independent >> bug. >> Because it's reproducable only with one specific font - maybe it's not an >> athens issue but bug in font itself and if someone could change/update >> Source Sans Pro to a newly version or to a similar font the problem will be >> solved? >> > > Did you try to use a new font version ? > I don't know exactly how embedded fonts are included. I tried to create > the font with a newer version but it did not work ( or at least I could > not see any difference). > > > >> >> Cheers, >> Alex >> > >

Oct. 8, 2015

Re: [Pharo-dev] [pharo-project/pharo-core] 450a4c: 50374

by Marcus Denker

That is oddâ¦ no changes integrated. I will reopen them. > On 08 Oct 2015, at 16:22, GitHub <noreply(a)github.com> wrote: > > Branch: refs/heads/5.0 > Home: https://github.com/pharo-project/pharo-core > Commit: 450a4cd448d62c4e6147098b5814662e5c5ad8e3 > https://github.com/pharo-project/pharo-core/commit/450a4cd448d62c4e6147098b… > Author: Jenkins Build Server <board(a)pharo-project.org> > Date: 2015-10-08 (Thu, 08 Oct 2015) > > Changed paths: > R ScriptLoader50.package/ScriptLoader.class/instance/pharo - scripts/script50373.st > A ScriptLoader50.package/ScriptLoader.class/instance/pharo - scripts/script50374.st > R ScriptLoader50.package/ScriptLoader.class/instance/pharo - updates/update50373.st > A ScriptLoader50.package/ScriptLoader.class/instance/pharo - updates/update50374.st > M ScriptLoader50.package/ScriptLoader.class/instance/public/commentForCurrentUpdate.st > > Log Message: > ----------- > 50374 > 16741 remove addModelMenuItemsTo:forMorph:hand: > https://pharo.fogbugz.com/f/cases/16741 > > 16729 Removing PluggableTextField from FT > https://pharo.fogbugz.com/f/cases/16729 > > 16744 ReflectiveMethod: make sure to not use DNU forwarding on invalidate > https://pharo.fogbugz.com/f/cases/16744 > > http://files.pharo.org/image/50/50374.zip > >

Oct. 8, 2015

[pharo-project/pharo-core]

by GitHub

Branch: refs/tags/50374 Home: https://github.com/pharo-project/pharo-core

Oct. 8, 2015

[pharo-project/pharo-core] 450a4c: 50374

by GitHub

Branch: refs/heads/5.0 Home: https://github.com/pharo-project/pharo-core Commit: 450a4cd448d62c4e6147098b5814662e5c5ad8e3 https://github.com/pharo-project/pharo-core/commit/450a4cd448d62c4e6147098b… Author: Jenkins Build Server <board(a)pharo-project.org> Date: 2015-10-08 (Thu, 08 Oct 2015) Changed paths: R ScriptLoader50.package/ScriptLoader.class/instance/pharo - scripts/script50373.st A ScriptLoader50.package/ScriptLoader.class/instance/pharo - scripts/script50374.st R ScriptLoader50.package/ScriptLoader.class/instance/pharo - updates/update50373.st A ScriptLoader50.package/ScriptLoader.class/instance/pharo - updates/update50374.st M ScriptLoader50.package/ScriptLoader.class/instance/public/commentForCurrentUpdate.st Log Message: ----------- 50374 16741 remove addModelMenuItemsTo:forMorph:hand: https://pharo.fogbugz.com/f/cases/16741 16729 Removing PluggableTextField from FT https://pharo.fogbugz.com/f/cases/16729 16744 ReflectiveMethod: make sure to not use DNU forwarding on invalidate https://pharo.fogbugz.com/f/cases/16744 http://files.pharo.org/image/50/50374.zip

Oct. 8, 2015