- Pharo-dev - lists.pharo.org

Re: [Pharo-project] [ANN] OpenGL specs parser/FFI methods generator

by Andreas Raab

On 4/13/2010 8:10 PM, Igor Stasenko wrote: > i just submitted a code, which could help us greatly in having > up-to-date and full OpenGL API coverage in Squeak. > > This package extracts data directly from spec files, used by OpenGL > folks for generating C headers (gl.h / glext.h etc). Sweet! I had no idea these spec files existed. Do you know if the spec files cover OpenGL ES as well? Also, have you run this against the existing OpenGL code to see if it produces matching output? (more to verify your translation than anything else) I'm definitely interested in looking at this. The original OpenGL calls were generated by code that I've lost in the mean time so it's great to have a way of recreating it from first principles! Cheers, - Andreas > So it is 100% correct. > Parsing these files is much easier than C headers. > There's also a lot of additional data , which helps to automatically > categorise the methods and generate correct callout code. > > The parsed data can be later used for generating an FFI/Alien OpenGL > callout code with full coverage of all existing extensions. > And you don't need to keep this package in image after you done > generating the code. > Or, it can stay, just make sure that you prune all parsed data by issuing: > > GLAPIData reset. > > The sources for parsing is located at: > > http://www.opengl.org/registry/#specfiles > > Place all *.spec and *.tm files into a single directory, and then issue: > > GLAPIData parseAll: (FileDirectory on: 'your/path'). > > > There is an example, how to generate an FFI callout methods. > > Use: > > GLAPIData gl generateFFIMethodsInto: (FileStream newFileNamed: > 'gl.st') forClass: 'OpenGL' > > Then, you can file-in the generated code right into your image: > > (FileStream oldFileNamed: 'gl.st') fileIn > > The resulting file is about 700kbytes big, having methods like: > > !OpenGL methodsFor: 'EXT_framebuffer_multisample' stamp: > 'Igor.Stasenko 4/14/2010 05:22'! > glRenderbufferStorageMultisampleEXT: target with: samples with: > internalformat with: width with: height > "This method was automatically generated from OpenGL specs" > "See http://squeaksource.com/OpenGLSpecs for details" > <apicall: void 'glRenderbufferStorageMultisampleEXT' (GLenum GLsizei > GLenum GLsizei GLsizei)> > ^ self externalCallFailed > > > P.S. I plan to use this data directly by native code, in a way like: > > glTexCoord2hNV: s with: t > <primitive: 'primitiveNativeCall' module: 'NativeBoostPlugin'> > ^ self glAPICall: 'glTexCoord2hNV' > > since i don't need a type data placed in method. I guess, Alien could > use similar approach. >

April 14, 2010

[Pharo-project] [ANN] OpenGL specs parser/FFI methods generator

by Igor Stasenko

Hello, i just submitted a code, which could help us greatly in having up-to-date and full OpenGL API coverage in Squeak. This package extracts data directly from spec files, used by OpenGL folks for generating C headers (gl.h / glext.h etc). So it is 100% correct. Parsing these files is much easier than C headers. There's also a lot of additional data , which helps to automatically categorise the methods and generate correct callout code. The parsed data can be later used for generating an FFI/Alien OpenGL callout code with full coverage of all existing extensions. And you don't need to keep this package in image after you done generating the code. Or, it can stay, just make sure that you prune all parsed data by issuing: GLAPIData reset. The sources for parsing is located at: http://www.opengl.org/registry/#specfiles Place all *.spec and *.tm files into a single directory, and then issue: GLAPIData parseAll: (FileDirectory on: 'your/path'). There is an example, how to generate an FFI callout methods. Use: GLAPIData gl generateFFIMethodsInto: (FileStream newFileNamed: 'gl.st') forClass: 'OpenGL' Then, you can file-in the generated code right into your image: (FileStream oldFileNamed: 'gl.st') fileIn The resulting file is about 700kbytes big, having methods like: !OpenGL methodsFor: 'EXT_framebuffer_multisample' stamp: 'Igor.Stasenko 4/14/2010 05:22'! glRenderbufferStorageMultisampleEXT: target with: samples with: internalformat with: width with: height "This method was automatically generated from OpenGL specs" "See http://squeaksource.com/OpenGLSpecs for details" <apicall: void 'glRenderbufferStorageMultisampleEXT' (GLenum GLsizei GLenum GLsizei GLsizei) > ^ self externalCallFailed P.S. I plan to use this data directly by native code, in a way like: glTexCoord2hNV: s with: t <primitive: 'primitiveNativeCall' module: 'NativeBoostPlugin'> ^ self glAPICall: 'glTexCoord2hNV' since i don't need a type data placed in method. I guess, Alien could use similar approach. -- Best regards, Igor Stasenko AKA sig.

April 14, 2010

Re: [Pharo-project] news about 1.0

by StÃ©phane Ducasse

thanks a lot for the information!. Stef On Apr 13, 2010, at 10:01 PM, laurent laffont wrote: > Hi, > > with last squeak-vm Linux binaries all Pharo tests are green and fonts working. So it may be a good candidate for the one click image. > > http://www.squeakvm.org/unix/release/Squeak-4.0.3.2202-linux_i386.tar.gz > > Cheers, > > Laurent Laffont > > > On Mon, Apr 12, 2010 at 11:40 AM, Henrik Johansen <henrik.s.johansen(a)veloxit.no> wrote: > > On Apr 11, 2010, at 11:26 03PM, StÃ©phane Ducasse wrote: > > > Hi guys > > > > mariano, marcus and adrian got some push to get 1.0 out. I was just on cc: > > I think that we made a mistake not to roll that publicly via the mailing-list (we will do that for 1.1). > > Now this is holidays time here, mariano is going to get married and adrian just moved so no > > internet connection. > > Tomorrow I will sync with marcus to know if I can help to get this 1.0 version out. > > If I remember the mails I saw, marcus was looking for the > > BitBltPluging binaries for the one click image on other platforms than mac os. > > > > Stef > Hum, aren't they usually built as internal? > In that case, the only reason for including plugin binaries would be if there's been bug fixes to it since the vm was shipped... > Which would probably mean you're providing a too old VM in the first place. > IIRC, On Windows, 3.11.8 / 4.0.2 would both include updated plugins, the main difference being 4.0.2 providing support for old image segment loading, but requiring in-image handling of quit events. > > Not sure about linux, since I don't use it. > Seems to be in a state of updating at the moment, with 5-6 commits over the weekend, so might be worth waiting for. > > Cheers, > Henry > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

April 13, 2010

Re: [Pharo-project] news about 1.0

by laurent laffont

Hi, with last squeak-vm Linux binaries all Pharo tests are green and fonts working. So it may be a good candidate for the one click image. http://www.squeakvm.org/unix/release/Squeak-4.0.3.2202-linux_i386.tar.gz Cheers, Laurent Laffont On Mon, Apr 12, 2010 at 11:40 AM, Henrik Johansen < henrik.s.johansen(a)veloxit.no> wrote: > > On Apr 11, 2010, at 11:26 03PM, StÃ©phane Ducasse wrote: > > > Hi guys > > > > mariano, marcus and adrian got some push to get 1.0 out. I was just on > cc: > > I think that we made a mistake not to roll that publicly via the > mailing-list (we will do that for 1.1). > > Now this is holidays time here, mariano is going to get married and > adrian just moved so no > > internet connection. > > Tomorrow I will sync with marcus to know if I can help to get this 1.0 > version out. > > If I remember the mails I saw, marcus was looking for the > > BitBltPluging binaries for the one click image on other platforms > than mac os. > > > > Stef > Hum, aren't they usually built as internal? > In that case, the only reason for including plugin binaries would be if > there's been bug fixes to it since the vm was shipped... > Which would probably mean you're providing a too old VM in the first place. > IIRC, On Windows, 3.11.8 / 4.0.2 would both include updated plugins, the > main difference being 4.0.2 providing support for old image segment loading, > but requiring in-image handling of quit events. > > Not sure about linux, since I don't use it. > Seems to be in a state of updating at the moment, with 5-6 commits over the > weekend, so might be worth waiting for. > > Cheers, > Henry > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >

April 13, 2010

Re: [Pharo-project] fixing underscores in a trivial manner thanks to SmalllintChecker and Lukas!

by Chris Muller

> Sure, that's why we version code. So I have to hunt around prior versions in the SCR just to discover who *really* wrote some method and when? Or, when debugging something and suspecting a method indicating recently "changed", I dig around the SCR, only to discover, "oh, actually Dan Ingalls' method really didn't change since 1999 afterall, it was just the assignment operator in 2010." Invalid accounting info all over the image. Sheesh, what a shame.. >> Just a friendly reminder that the FixUnderscores package is available >> to do this with preservation.. > > It's broken, it only works 90%. I've been using it for years with no problem. What do you mean "broken?"

April 13, 2010

Re: [Pharo-project] About getting HashedCollection (another dream happening)

by Levente Uzonyi

On Tue, 13 Apr 2010, Chris Muller wrote: > Peter mentioned VA, whose Set's swallow nil's. Perhaps that's the missing link in the communication here. Igor is talking about *Squeak*, which does not swallow nil's, it throws an error. So, to "fix consequences instead of causes" in the case of Squeak, the software would have to be doing something like: [ #(1 2 3 nil) asSet ] on: Error do: [ : err | "ignore it!" ] which, deserves to break if it exists at all.. Squeak doesn't raise an error, but adds nil to the set (since 4.1), so there's no need for such tricks. If you want to ignore nils, just #reject: or #remove: them. Levente On Mon, Apr 12, 2010 at 10:33 AM, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: > igor > > peter snippets was just a dull example. Peter mentioned that some people fix consequences instead of causes.... > > > Stef > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > _______________________________________________ Pharo-project mailing list Pharo-project(a)lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

April 13, 2010

Re: [Pharo-project] fixing underscores in a trivial manner thanks to SmalllintChecker and Lukas!

by Lukas Renggli

> Unfortunately, the valuable accounting information is wiped out by > this approach isn't it? Sure, that's why we version code. > Just a friendly reminder that the FixUnderscores package is available > to do this with preservation.. It's broken, it only works 90%. Lukas -- Lukas Renggli www.lukas-renggli.ch

April 13, 2010

Re: [Pharo-project] About getting HashedCollection (another dream happening)

by Chris Muller

Peter mentioned VA, whose Set's swallow nil's. Perhaps that's the missing link in the communication here. Igor is talking about *Squeak*, which does not swallow nil's, it throws an error. So, to "fix consequences instead of causes" in the case of Squeak, the software would have to be doing something like: [ #(1 2 3 nil) asSet ] on: Error do: [ : err | "ignore it!" ] which, deserves to break if it exists at all.. On Mon, Apr 12, 2010 at 10:33 AM, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: > igor > > peter snippets was just a dull example. Peter mentioned that some people fix consequences instead of causes.... > > > Stef > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >

April 13, 2010

Re: [Pharo-project] fixing underscores in a trivial manner thanks to SmalllintChecker and Lukas!

by Chris Muller

Unfortunately, the valuable accounting information is wiped out by this approach isn't it? Yes, RB's Rewrite tool is great, but for this job, due deference should be given to the original history and authorship, because our politics of the "assignment syntax" is not meaningful to the original software. Just a friendly reminder that the FixUnderscores package is available to do this with preservation.. On Sun, Apr 11, 2010 at 5:51 PM, Fernando olivero <oliverof(a)lu.unisi.ch> wrote: > Hi all, i just wanted to re post the fix underscore code Lukas posted a while ago. > > I've just imported Connectors package into pharo, (i'm trying to get a minimal version working ), and stumbled upon the problem of the underscores.. > but now thanks to SmallInt rules it was easy and fast to overcome this problem! > > Thanks!!!! > > Fernando > > > pd: From previous mail from lukas. > > > 1. Load the code: > > Â Gofer new > Â Â Â squeaksource: 'rb'; > Â Â Â package: 'AST-Core'; > Â Â Â package: 'Refactoring-Core'; > Â Â Â load. > > 2. Select the code (packages) you want to fix: > > Â environment := BrowserEnvironment new > Â Â Â forPackageNames: #('Connectors-Base' 'Connectors-Lines and Curves'). > > 3. Create the transformation rule: > > Â rule := RBUnderscoreAssignmentRule new. > > 4. Perform the search: > > Â SmalllintChecker runRule: rule onEnvironment: environment. > > 5. Perform the transformation: > > Â change := CompositeRefactoryChange new. > Â change changes: rule changes. > Â change execute > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >

April 13, 2010

Re: [Pharo-project] A port of Pharo: Redline Smalltalk ...

by StÃ©phane Ducasse

> These would be migrated from Squeak to Pharo, right? Yes but not just that, the idea is to have a look if the idea is good (it is good for 2171) and if it applies to pharo. Sometimes there is a change in squeak but you want it differently in pharo. So this is a nice way to learn. For the tests this is also the occasion to check the tests we have and check if they can be improve, if the tests in squeak make sense in pharo... > On Sun, Apr 11, 2010 at 10:01 PM, StÃ©phane Ducasse > <stephane.ducasse(a)inria.fr> wrote: >> Richard >> >> Nicolas checked systematically that add: method returns their argument. >> >> - Issue 2171: add consistency in MC, Exception and other >> >> for the tests: >> this a list of possible tests that would be worth to migrate to pharo. >> - Issue 2238: tests for FileStream behavior >> - Issue 1920: Better PureBehaviorTests >> - Issue 1934: some tests for WeakRegistry >> >> Let me know if you need advices. >> >> Stef >> >> On Apr 11, 2010, at 6:16 PM, Richard Durr wrote: >> >>> I'd like to contribute, but I am not good enough for the difficult >>> ones to come to pleasing solutions, and I somehow don't feel the other >>> ones accessable for me :/ >>> >>> On Sat, Apr 10, 2010 at 9:21 PM, StÃ©phane Ducasse >>> <stephane.ducasse(a)inria.fr> wrote: >>>> Hi richard >>>> >>>> some changes are changes coming from squeak and in such a case there is the squeak change >>>> It does not mean that we can integrate the change as it is. >>>> Thanks for looking at fixes and bug entries >>>> >>>> Stef >>>> >>>> On Apr 10, 2010, at 7:52 PM, Richard Durr wrote: >>>> >>>>> I am still trying to grok how this issue-fixing works. For example, in >>>>> "Difficulty:Easy" there are some issues which consist mostly of >>>>> ChangeSets of something. O_o >>>> >>>> O_o >>>> >>>> Nice :) >>>> >>>> >>>>> >>>>> On Tue, Apr 6, 2010 at 10:50 AM, StÃ©phane Ducasse >>>>> <stephane.ducasse(a)inria.fr> wrote: >>>>>> Yes this is the vision we have. >>>>>> Now if people would help for simple things in pharo we would have more time for such issues. >>>>>> We got several people doing PhD on modules and related. We studied namespace and other kind of >>>>>> points so we have an idea of what could be a solution but it should be implemented >>>>>> tested..... >>>>>> >>>>>> Stef >>>>>> >>>>>> >>>>>> On Apr 6, 2010, at 1:14 AM, Richard Durr wrote: >>>>>> >>>>>>> No, it's not. I would be far more valuable to have modularized images >>>>>>> though: Just like Next/Apples Nibs or Bundle-mechanism. These >>>>>>> image-modules would be edited from a development-image-bundle. They >>>>>>> could be used to form stand-alone applications already clean of the >>>>>>> development code or libraries that contain only a diff to the base >>>>>>> classes, these library-images could then be combined to form more >>>>>>> complex images. They could also contain their own namespaceâ¦ and they >>>>>>> would replace the non-modelling packaging mechanism, but I digressâ¦ ^^ >>>>>>> >>>>>>> | packageWorkedOn | >>>>>>> packageWorkedOn := Browser newPackageNamed: 'Calculator'. >>>>>>> packageWorkedOn importPackage: 'Seaside'. >>>>>>> "Create code in the Browserâ¦" >>>>>>> packageWorkedOn saveAsMonticello. >>>>>>> packageWordedOn saveAsPackage. >>>>>>> packageWorkedOn saveAsApplication. >>>>>>> >>>>>>> or something like that â¦ ^^ >>>>>>> >>>>>>> On Mon, Apr 5, 2010 at 9:08 PM, nullPointer <epicfan(a)gmail.com> wrote: >>>>>>>> >>>>>>>> jarober says: >>>>>>>> >>>>>>>>> The reason these "Smalltalk for .NET" and "Smalltalk for the JVM" >>>>>>>>> projects never seem to come off is simple - Smalltalk isn't just flat >>>>>>>>> text in an editor. Smalltalk is the entire interactive environment. >>>>>>>>> It would be fairly simple to get a syntax parser, but it wouldn't be >>>>>>>>> Smalltalk. It would be Ruby or Python with Smalltalk syntax. >>>>>>>>> Somewhat useful perhaps, but not really Smalltalk. >>>>>>>> >>>>>>>> :( >>>>>>>> >>>>>>>> Hopefully it would have sufficient knowledge to be able to do something >>>>>>>> thus. >>>>>>>> >>>>>>>> Sorry, but I believe that it is the way that had to take commercial >>>>>>>> Smalltalk. It surprises to me that Cincom and company do not see it. >>>>>>>> >>>>>>>> Regards. >>>>>>>> -- >>>>>>>> View this message in context: http://n4.nabble.com/A-port-of-Pharo-Redline-Smalltalk-tp1751884p1751963.ht… >>>>>>>> Sent from the Pharo Smalltalk mailing list archive at Nabble.com. >>>>>>>> >>>>>>>> _______________________________________________ >>>>>>>> Pharo-project mailing list >>>>>>>> Pharo-project(a)lists.gforge.inria.fr >>>>>>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >>>>>>>> >>>>>>> >>>>>>> _______________________________________________ >>>>>>> Pharo-project mailing list >>>>>>> Pharo-project(a)lists.gforge.inria.fr >>>>>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >>>>>> >>>>>> >>>>>> _______________________________________________ >>>>>> Pharo-project mailing list >>>>>> Pharo-project(a)lists.gforge.inria.fr >>>>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >>>>>> >>>>> >>>>> _______________________________________________ >>>>> Pharo-project mailing list >>>>> Pharo-project(a)lists.gforge.inria.fr >>>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >>>> >>>> >>>> _______________________________________________ >>>> Pharo-project mailing list >>>> Pharo-project(a)lists.gforge.inria.fr >>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >>>> >>> >>> _______________________________________________ >>> Pharo-project mailing list >>> Pharo-project(a)lists.gforge.inria.fr >>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >> >> >> _______________________________________________ >> Pharo-project mailing list >> Pharo-project(a)lists.gforge.inria.fr >> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >> > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project

April 13, 2010