May 2014 - Pharo-dev - lists.pharo.org

Re: [Pharo-dev] Starting making bindings for OpenCL using NativeBoost.

by Alexandre Bergel

Excellent!!!!! Go go go! Alexandre On May 8, 2014, at 2:32 AM, Ronie Salgado <roniesalg(a)gmail.com> wrote: > Hello, > > I started working in making some basic bindings for OpenCL using NativeBoost. > > Currently I can run basics OpenCL OpenCL kernels. I tried to generate quickly some procedural images which was a success. > > I am going to be showing some more cool stuffs when mixing OpenCL+OpenGL+SDL2. > > Currently uploading my work into http://smalltalkhub.com/#!/~ronsaldo/OpenCL . No configuration or stable release yet. > > Greetings, > Ronie > <opencl-screenshot.png> -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

May 8, 2014

Re: [Pharo-dev] WTF and Programming Languages

by Sven Van Caekenberghe

On 08 May 2014, at 10:58, Nicolai Hess <nicolaihess(a)web.de> wrote: > 2014-05-08 10:39 GMT+02:00 phil(a)highoctane.be <phil(a)highoctane.be>: > On Thu, May 8, 2014 at 9:37 AM, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > > On 07 May 2014, at 23:25, Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com> wrote: > > > I love these kind of snippets :) > > > > I have built a collection of freak expressions in Squeak3.9. > > Pharo 3.0 is less spectacular, above collection is shrinking dangerously... > > but there are still some not so unsurprising features to test for: > > These are actually not that funny, we should try to do something about them, or at least discuss them a bit. > > > 'hello' copyReplaceFrom: 6 to: 4 with: 'world'. > > #copyReplaceFrom:to:with: seems overloaded with the insert case, is that really necessary ? > still the fact that the $o appears twice feels like a bug, no ? > > 'Helloworldo' was indeed very surprising. Silent failures like this seem pretty dangerous to me. > > Are you sure this is a failure ? :) > From the comment of SequenceableCollection>>copyReplaceFrom: start to: stop with: replacementCollection > > ... If stop is less than start, then this is an insertion;....start = size+1 means append after last character.... The why is the result not 'Helloworld' ? or 'Hellworldo' ? Why does the $o appear twice ?

May 8, 2014

Re: [Pharo-dev] WTF and Programming Languages

by Nicolai Hess

2014-05-08 10:39 GMT+02:00 phil(a)highoctane.be <phil(a)highoctane.be>: > On Thu, May 8, 2014 at 9:37 AM, Sven Van Caekenberghe <sven(a)stfx.eu>wrote: > >> >> On 07 May 2014, at 23:25, Nicolas Cellier < >> nicolas.cellier.aka.nice(a)gmail.com> wrote: >> >> > I love these kind of snippets :) >> > >> > I have built a collection of freak expressions in Squeak3.9. >> > Pharo 3.0 is less spectacular, above collection is shrinking >> dangerously... >> > but there are still some not so unsurprising features to test for: >> >> These are actually not that funny, we should try to do something about >> them, or at least discuss them a bit. >> >> > 'hello' copyReplaceFrom: 6 to: 4 with: 'world'. >> >> #copyReplaceFrom:to:with: seems overloaded with the insert case, is that >> really necessary ? >> still the fact that the $o appears twice feels like a bug, no ? >> > > 'Helloworldo' was indeed very surprising. Silent failures like this seem > pretty dangerous to me. > Are you sure this is a failure ? :) >From the comment of SequenceableCollection>>copyReplaceFrom: start to: stop with: replacementCollection ... If stop is less than start, then this is an insertion;....start = size+1 means append after last character....

May 8, 2014

Re: [Pharo-dev] WTF and Programming Languages

by Sven Van Caekenberghe

On 08 May 2014, at 10:39, phil(a)highoctane.be wrote: > The second - 2. I do not get as it prints a -2 SmallInteger. What's wrong there? Do you mean - 2 shouldn't be parsable? That I can get. It is an error (the space). It should either be -2 or 2 negated but - 2 is not a message send ...

May 8, 2014

Re: [Pharo-dev] WTF and Programming Languages

by Sven Van Caekenberghe

On 08 May 2014, at 10:39, phil(a)highoctane.be wrote: > > '' , #[119 104 97 116 63]. > > #[] , 'what?'. "honour is safe, it's sorta symmetric..." > > that is the ByteString/ByteArray dichotomy at work, may enforce the #species of the receiver and argument to be equal ? > > Would you mind enlighten me on this dichotomy? (String new: 10 withAll: $a) replaceFrom: 1 to: 3 with: #[65 66 67] startingAt: 1. (ByteArray new: 10 withAll: 16) replaceFrom: 1 to: 3 with: 'ABC' startingAt: 1. Primitive 105 allows you to mix ByteStrings and ByteArrays which is crucial for making certain system operations fast. Of course, once you take real encoding into account, this breaks.

May 8, 2014

Re: [Pharo-dev] WTF and Programming Languages

by phil＠highoctane.be

On Thu, May 8, 2014 at 9:37 AM, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > > On 07 May 2014, at 23:25, Nicolas Cellier < > nicolas.cellier.aka.nice(a)gmail.com> wrote: > > > I love these kind of snippets :) > > > > I have built a collection of freak expressions in Squeak3.9. > > Pharo 3.0 is less spectacular, above collection is shrinking > dangerously... > > but there are still some not so unsurprising features to test for: > > These are actually not that funny, we should try to do something about > them, or at least discuss them a bit. > > > 'hello' copyReplaceFrom: 6 to: 4 with: 'world'. > > #copyReplaceFrom:to:with: seems overloaded with the insert case, is that > really necessary ? > still the fact that the $o appears twice feels like a bug, no ? > 'Helloworldo' was indeed very surprising. Silent failures like this seem pretty dangerous to me. > > > (5 to: 10) at: 1.3. > > understandable, scary, wrong, maybe use #asInteger on the index ? > Especially when the index is intention revealing as in 'anInteger' :-( > > > '' , #[119 104 97 116 63]. > > #[] , 'what?'. "honour is safe, it's sorta symmetric..." > > that is the ByteString/ByteArray dichotomy at work, may enforce the > #species of the receiver and argument to be equal ? > Would you mind enlighten me on this dichotomy? In ByteArray I do see: asString "Convert to a String with Characters for each byte. Fast code uses primitive that avoids character conversion" ^ (String new: self size) replaceFrom: 1 to: self size with: self Ah, replaceFrom:to:with: again... It is the same Collection code for all of these. And now, we get >>asString, >>asWideString And in String>>new: sizeRequested "Answer an instance of this class with the number of indexable variables specified by the argument, sizeRequested." self == String ifTrue:[^ByteString new: sizeRequested] ifFalse:[^self basicNew: sizeRequested]. Mmh, okay... 1) no primitive in a method that says "Fast code uses primitive to avoid character conversion": is that normal? 2) self == String, so ^ByteString new I'd love to know if there is a clear explanation of all of this somewhere. Is there one? > > "Cases of special syntax. Is the behavior officially supported in some > EBNF?" > > -2r-1. > > - 2. "Note the space between the two letters" > > These are at the (number) syntax level. Again scary, we should be (much) > stricter in parsing. > The second - 2. I do not get as it prints a -2 SmallInteger. What's wrong there? Do you mean - 2 shouldn't be parsable? That I can get. Phil > > > 2014-05-07 22:47 GMT+02:00 phil(a)highoctane.be <phil(a)highoctane.be>: > > > > Le 7 mai 2014 22:17, "kilon alios" <kilon.alios(a)gmail.com> a Ã©crit : > > > > > > > > > > I dont know if you guys are aware of this talk, very funny , quite > weird , so I thought why not share it with you > > > > > > https://www.destroyallsoftware.com/talks/wat > > > > Yeah, Amber kind of alleviates some of these pain points... > > > > >

May 8, 2014

Re: [Pharo-dev] WTF and Programming Languages

by Yuriy Tymchuk

On 08 May 2014, at 09:37, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > > On 07 May 2014, at 23:25, Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com> wrote: > >> I love these kind of snippets :) >> >> I have built a collection of freak expressions in Squeak3.9. >> Pharo 3.0 is less spectacular, above collection is shrinking dangerously... >> but there are still some not so unsurprising features to test for: > > These are actually not that funny, we should try to do something about them, or at least discuss them a bit. > >> 'hello' copyReplaceFrom: 6 to: 4 with: 'world'. > > #copyReplaceFrom:to:with: seems overloaded with the insert case, is that really necessary ? > still the fact that the $o appears twice feels like a bug, no ? > >> (5 to: 10) at: 1.3. Amazing. But this is even better (3 to: 8 by: 2) at: 1.3 > > understandable, scary, wrong, maybe use #asInteger on the index ? > >> '' , #[119 104 97 116 63]. >> #[] , 'what?'. "honour is safe, it's sorta symmetric..." > > that is the ByteString/ByteArray dichotomy at work, may enforce the #species of the receiver and argument to be equal ? > >> "Cases of special syntax. Is the behavior officially supported in some EBNF?" >> -2r-1. >> - 2. "Note the space between the two letters" > > These are at the (number) syntax level. Again scary, we should be (much) stricter in parsing. > >> 2014-05-07 22:47 GMT+02:00 phil(a)highoctane.be <phil(a)highoctane.be>: >> >> Le 7 mai 2014 22:17, "kilon alios" <kilon.alios(a)gmail.com> a Ã©crit : >> >> >>> >>> I dont know if you guys are aware of this talk, very funny , quite weird , so I thought why not share it with you >>> >>> https://www.destroyallsoftware.com/talks/wat >> >> Yeah, Amber kind of alleviates some of these pain points... >> > >

May 8, 2014

Re: [Pharo-dev] Starting making bindings for OpenCL using NativeBoost.

by kilon alios

wow nice, opencl can seriously speed things up for many diffirent kind's of calculation. Blender <http://www.blender.org/> is actually porting OpenCL for its rendering engine cycles. Cycles already uses CUDA. They say that openCL is quite buggy yet and not well supported. Whats your experience with it ? In Blender CUDA alone speed things up 3 times compared to my CPU. For serious GPUs this number can go easily up to 10 times. I have even heard that in some cases the speed up reach 100 times. Are you going to expose also an API for non graphic stuff ? On Thu, May 8, 2014 at 9:32 AM, Ronie Salgado <roniesalg(a)gmail.com> wrote: > Hello, > > I started working in making some basic bindings for OpenCL using > NativeBoost. > > Currently I can run basics OpenCL OpenCL kernels. I tried to generate > quickly some procedural images which was a success. > > I am going to be showing some more cool stuffs when mixing > OpenCL+OpenGL+SDL2. > > Currently uploading my work into > http://smalltalkhub.com/#!/~ronsaldo/OpenCL . No configuration or stable > release yet. > > Greetings, > Ronie >

May 8, 2014

Re: [Pharo-dev] WTF and Programming Languages

by Sven Van Caekenberghe

On 07 May 2014, at 23:25, Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com> wrote: > I love these kind of snippets :) > > I have built a collection of freak expressions in Squeak3.9. > Pharo 3.0 is less spectacular, above collection is shrinking dangerously... > but there are still some not so unsurprising features to test for: These are actually not that funny, we should try to do something about them, or at least discuss them a bit. > 'hello' copyReplaceFrom: 6 to: 4 with: 'world'. #copyReplaceFrom:to:with: seems overloaded with the insert case, is that really necessary ? still the fact that the $o appears twice feels like a bug, no ? > (5 to: 10) at: 1.3. understandable, scary, wrong, maybe use #asInteger on the index ? > '' , #[119 104 97 116 63]. > #[] , 'what?'. "honour is safe, it's sorta symmetric..." that is the ByteString/ByteArray dichotomy at work, may enforce the #species of the receiver and argument to be equal ? > "Cases of special syntax. Is the behavior officially supported in some EBNF?" > -2r-1. > - 2. "Note the space between the two letters" These are at the (number) syntax level. Again scary, we should be (much) stricter in parsing. > 2014-05-07 22:47 GMT+02:00 phil(a)highoctane.be <phil(a)highoctane.be>: > > Le 7 mai 2014 22:17, "kilon alios" <kilon.alios(a)gmail.com> a Ã©crit : > > > > > > I dont know if you guys are aware of this talk, very funny , quite weird , so I thought why not share it with you > > > > https://www.destroyallsoftware.com/talks/wat > > Yeah, Amber kind of alleviates some of these pain points... >

May 8, 2014

Pharo 4.0 Inbox

by Sean P. DeNigris

I created http://smalltalkhub.com/mc/Pharo/Pharo40Inbox/main and made it public writable ----- Cheers, Sean -- View this message in context: http://forum.world.st/Pharo-4-0-Inbox-tp4758294.html Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

May 8, 2014