- Pharo-dev - lists.pharo.org

Re: [Pharo-dev] Athens: issue with clipping and/or path drawing

by Yahoo-Plantec

Thanks Nicolai! it works with the last vm. Thanks a lot. Cheers Alain On 13 oct. 2014, at 09:05, Nicolai Hess <nicolaihess(a)web.de> wrote: > 2014-10-13 8:53 GMT+02:00 Alain Plantec <alain.plantec(a)yahoo.com>: > Dear All, > The border is not well rendered when a submorph is partly > outside the bounds of the AthensWrapMorph. > The AthensWrapMorph is set to clip its submorphs. > See the two small screenshots with an ellipse morph. > Is it a known issue or am I doing something wrong ? > > | a e | > a := AthensWrapMorph new clipSubmorphs: true. > e := EllipseMorph new. > e position: 100@100. > a addMorph: e. > a openInWorld > > Thanks > Alain > > <Screen Shot 2014-10-13 at 08.42.18.png><Screen Shot 2014-10-13 at 08.42.33.png> > > > There was an issue on windows and mac os. The vm didn't use the latest > cairo lib: > 13364 > Mac OS X and Windows VM with newer Cairo > > Does it work for you with the latest pharo-vm? > > Nicolai > >

Oct. 13, 2014

Re: [Pharo-dev] Athens: issue with clipping and/or path drawing

by Nicolai Hess

2014-10-13 8:53 GMT+02:00 Alain Plantec <alain.plantec(a)yahoo.com>: > Dear All, > The border is not well rendered when a submorph is partly > outside the bounds of the AthensWrapMorph. > The AthensWrapMorph is set to clip its submorphs. > See the two small screenshots with an ellipse morph. > Is it a known issue or am I doing something wrong ? > > | a e | > a := AthensWrapMorph new clipSubmorphs: true. > e := EllipseMorph new. > e position: 100@100. > a addMorph: e. > a openInWorld > > Thanks > Alain > > There was an issue on windows and mac os. The vm didn't use the latest cairo lib: 13364 <https://pharo.fogbugz.com/default.asp?13364> Mac OS X and Windows VM with newer Cairo Does it work for you with the latest pharo-vm? Nicolai

Oct. 13, 2014

Athens: issue with clipping and/or path drawing

by Alain Plantec

Dear All, The border is not well rendered when a submorph is partly outside the bounds of the AthensWrapMorph. The AthensWrapMorph is set to clip its submorphs. See the two small screenshots with an ellipse morph. Is it a known issue or am I doing something wrong ? | a e | a := AthensWrapMorph new clipSubmorphs: true. e := EllipseMorph new. e position: 100@100. a addMorph: e. a openInWorld Thanks Alain

Oct. 13, 2014

Re: [Pharo-dev] Working configuration of OS and Graphiccard for Woden

by Nicolai Hess

2014-09-28 23:58 GMT+02:00 Nicolai Hess <nicolaihess(a)web.de>: > Is Woden supposed to work on windows? > I tried > Gofer new smalltalkhubUser: 'ronsaldo' project: 'Woden'; package: > 'ConfigurationOfWoden'; load. (Smalltalk at: #ConfigurationOfWoden) > loadBleedingEdge > on windows but it failed (some NB error). > I just want to know if anyone tried woden on Windows already. > I tried this on linux too, but it fails. > (But I wasn't able to run Roassal3D or NBOpenGL on my > linux box so I hadn't high expectations, > but at least on my windows box I could get NBOpenGL > working) > > > Nicolai > solved: It was a driver issue. With the latest ATI driver on windows woden examples are working.

Oct. 12, 2014

Re: [Pharo-dev] OpenGL experts needed

by Nicolai Hess

Thank you Ronie, 2014-10-12 21:47 GMT+02:00 Ronie Salgado <roniesalg(a)gmail.com>: > Hello, > > You should start trying to use Woden and Woden-Roassal. Everything of > Roassal 3D will be moved to Woden. > http://smalltalkhub.com/#!/~ronsaldo/Woden . To install it you only have > to run this script: > > Gofer new smalltalkhubUser: 'ronsaldo' project: 'Woden'; package: > 'ConfigurationOfWoden'; load. (Smalltalk at: #ConfigurationOfWoden) > loadBleedingEdgeIn Woden I have many bugs to fix, which I havent because I > have been busy the last month making the new FFI. > > I have still problems with NBOpenGL with an ATI Radeon HD 4250 >> Graphic card > > What graphic driver do you have? AMD stopped providing support to the > graphics cards older than the HD 5xxx, If i remember it correctly. I guess > that you are using the open source drivers. > I feel so stupid. On windows I was pretty sure it is the most recent one. (I checked the driver settings and on "update" it shows "Already up to date"-message). Now I manually downloaded the latest driver from ATI and ... ... It works now! (and the woden examples are working too). Nicolai > > Greetings, > Ronie > > 2014-10-12 14:03 GMT-03:00 Nicolai Hess <nicolaihess(a)web.de>: > > like described here >> http://forum.world.st/Roassal-3D-empty-windows-td4749249.html >> and here >> http://forum.world.st/NBOpenGL-on-Windows-td4749309.html >> >> I have still problems with NBOpenGL with an ATI Radeon HD 4250 >> Graphic card >> >> No way to get it to work on linux (dont' know why). >> >> On windows I had the issue that the rendering only happens in >> the upper window corner (rendering only happens within the >> initial morph/framebuffer bounds). >> >> I finally solved this issue by reattaching the renderbuffers after every >> resize. >> ( in NBGLFrameBuffer>>#resize: >> I call >> self bind >> and >> for all attachments I call #resize: and #attachTo:as: >> (I had to change the attachments set to store the attachment type as >> well)) >> >> Now my question is, is it a bug in ATIs opengl implementation >> Or are all other users just lucky, that it works for them? >> Or it is just uncommon to work with framebuffer objects this way? >> Don't know if my solution just works now, but is otherwise wrong >> or resuls in leaks or other side effects. >> >> (Actually there are tons of examples and tutorials online about >> framebuffer objects and render-to-texture, but none I found covers the >> case >> of a changed framebuffer size) >> >> Anyone has some expert knowledge on using framebuffers for >> render-to-texture >> where this target texture may change its size? >> Anyone working with NBOpenGL and hasn't this issue can tell me >> which Grahic card he uses? >> Or anyone knows something about bugs in ATIs OpenGL driver which >> may be responsible for this behavior? >> (I read about another bug with render-to-texture but that one only >> happened for NVIDIA cards) >> >> thanks in advance >> Nicolai >> >> (btw in NBGLCurveRenderer there is still a problem with >> NBGLFragmentProgramARB vs NBGLFragmentProgram. >> All this with Pharo3.0 Latest update: #30858 latest NBOpenGL image >> from Pharo Contribution Jenkins) >> >> >> >> >

Oct. 12, 2014

Re: [Pharo-dev] Serious problem with ConfigurationOfBootstrap - Error: Apparent loop in import expansion

by phil＠highoctane.be

It turned out that the problem appeared randomly on various configurations and the cause is possibly a low memory condition. Also possible is an instable smalltalkhub. The problems were blocking us for a while with no clear cause. That is quite annoying. Phil Le 12 oct. 2014 21:35, "Torsten Bergmann" <astares(a)gmx.de> a Ã©crit : > > Hi Phil, > > nothing I can reproduce here. I tried with the latest/fresh > Pharo 3.0 (build 30858) image. No problem to load from script or config > browser. > Also CI is green. > > Are you already using Pharo 4.0? Then it might be related that > the prerequisite Seaside version may be not yet be ready/supported > for Pharo 4.0. > https://ci.inria.fr/pharo-contribution/job/Pharo4Seaside/ > > Bye > T. > > > > > >

Oct. 12, 2014

Re: [Pharo-dev] OpenGL experts needed

by Ronie Salgado

Hello, You should start trying to use Woden and Woden-Roassal. Everything of Roassal 3D will be moved to Woden. http://smalltalkhub.com/#!/~ronsaldo/Woden . To install it you only have to run this script: Gofer new smalltalkhubUser: 'ronsaldo' project: 'Woden'; package: 'ConfigurationOfWoden'; load. (Smalltalk at: #ConfigurationOfWoden) loadBleedingEdge In Woden I have many bugs to fix, which I havent because I have been busy the last month making the new FFI. I have still problems with NBOpenGL with an ATI Radeon HD 4250 > Graphic card What graphic driver do you have? AMD stopped providing support to the graphics cards older than the HD 5xxx, If i remember it correctly. I guess that you are using the open source drivers. Greetings, Ronie 2014-10-12 14:03 GMT-03:00 Nicolai Hess <nicolaihess(a)web.de>: > like described here > http://forum.world.st/Roassal-3D-empty-windows-td4749249.html > and here > http://forum.world.st/NBOpenGL-on-Windows-td4749309.html > > I have still problems with NBOpenGL with an ATI Radeon HD 4250 > Graphic card > > No way to get it to work on linux (dont' know why). > > On windows I had the issue that the rendering only happens in > the upper window corner (rendering only happens within the > initial morph/framebuffer bounds). > > I finally solved this issue by reattaching the renderbuffers after every > resize. > ( in NBGLFrameBuffer>>#resize: > I call > self bind > and > for all attachments I call #resize: and #attachTo:as: > (I had to change the attachments set to store the attachment type as well)) > > Now my question is, is it a bug in ATIs opengl implementation > Or are all other users just lucky, that it works for them? > Or it is just uncommon to work with framebuffer objects this way? > Don't know if my solution just works now, but is otherwise wrong > or resuls in leaks or other side effects. > > (Actually there are tons of examples and tutorials online about > framebuffer objects and render-to-texture, but none I found covers the case > of a changed framebuffer size) > > Anyone has some expert knowledge on using framebuffers for > render-to-texture > where this target texture may change its size? > Anyone working with NBOpenGL and hasn't this issue can tell me > which Grahic card he uses? > Or anyone knows something about bugs in ATIs OpenGL driver which > may be responsible for this behavior? > (I read about another bug with render-to-texture but that one only > happened for NVIDIA cards) > > thanks in advance > Nicolai > > (btw in NBGLCurveRenderer there is still a problem with > NBGLFragmentProgramARB vs NBGLFragmentProgram. > All this with Pharo3.0 Latest update: #30858 latest NBOpenGL image > from Pharo Contribution Jenkins) > > > >

Oct. 12, 2014

Re: [Pharo-dev] Serious problem with ConfigurationOfBootstrap - Error: Apparent loop in import expansion

by Torsten Bergmann

Hi Phil, Â nothing I can reproduce here. I tried with the latest/fresh Pharo 3.0 (build 30858) image. No problem to load from script or config browser. Also CI is green. Are you already using Pharo 4.0? Then it might be related that the prerequisite Seaside version may be not yet be ready/supported for Pharo 4.0. https://ci.inria.fr/pharo-contribution/job/Pharo4Seaside/ Bye T.

Oct. 12, 2014

Re: [Pharo-dev] Artefact documentation is now in english

by phil＠highoctane.be

404 Not Found. Something wrong with the build. Phil On Sun, Oct 12, 2014 at 8:44 PM, Tudor Girba <tudor(a)tudorgirba.com> wrote: > Thanks! > > Doru > > On Sun, Oct 12, 2014 at 7:36 PM, stepharo <stepharo(a)free.fr> wrote: > >> Hi guys >> >> I finished the translation of the chapter on Artefact the pdf generator >> library developed by O. Auverlot and G. Larcheveque. >> >> https://ci.inria.fr/pharo-contribution/job/PharoForTheEnterprise/ >> lastSuccessfulBuild/artifact/Artefact/Artefact.pillar.pdf >> >> Stef >> >> > > > -- > www.tudorgirba.com > > "Every thing has its own flow" >

Oct. 12, 2014

Re: [Pharo-dev] Artefact documentation is now in english

by Tudor Girba

Thanks! Doru On Sun, Oct 12, 2014 at 7:36 PM, stepharo <stepharo(a)free.fr> wrote: > Hi guys > > I finished the translation of the chapter on Artefact the pdf generator > library developed by O. Auverlot and G. Larcheveque. > > https://ci.inria.fr/pharo-contribution/job/PharoForTheEnterprise/ > lastSuccessfulBuild/artifact/Artefact/Artefact.pillar.pdf > > Stef > > -- www.tudorgirba.com "Every thing has its own flow"

Oct. 12, 2014