- Pharo-dev - lists.pharo.org

Re: [Pharo-dev] Roadmap on tools?

by Marcus Denker

On 26 Aug 2014, at 10:03, Serge Stinckwich <serge.stinckwich(a)gmail.com> wrote: > On Mon, Aug 25, 2014 at 11:17 PM, Alain Rastoul <alr.dev(a)free.fr> wrote: >> +1 for the GTInspector, and the Moose tools/paradigm too (Roassal, Glamour, >> Moose and others), all that stuff is great step forward and could arouse >> interest from doubtful people. >> I remember myself failing to show some collegues at work how the smalltalk >> system could be a cool tool to play with, even for people sticking on >> dotNet, Delphi or C++. >> And sometimes they remember that too ... >> (Smalltalk? Squeak? -at that time- that blinking and poping toy ? hahaha >> ...) >> :( >> Still working on that like a flea (?- a morpion) >> >> Morphic removed is good news - clumsy, buggy and weird - but I don't >> understand the relationship with GTInspector ? >> I googled about that and just found a post of you about Bloc in the mailing >> list, it sounds like a good idea, and I'm sure you'll manage to do it >> cleanly, but I'm also very curious about that: big bang or dependency >> injection and small steps? other patterns, techniques ? a link on Bloc ? >> I'm also curious about Spec and it's status after it's change to GPL ? Will >> it be supported in the future ? What are the alternatives ? > > Yes, apparently spec is distributed now under a dual licence : MIT > when used as an external library (not sure what it means when you use > Smalltalk) > and GPL when integrated in an IDE ... I think that this is a potential > problem for Pharo. > GPL is not compatible with Pharo. All code that is part of the Pharo main distribution is either historical (Apple Licence) or MIT. We even let people sign a document that makes this clear. New code has to be MIT, we do not accept any other license (as part of the main distribution). e.g. Zinc was done because the HTTP server we were using was made GPL (it did not have a licence when we started to use it). Marcus

Aug. 26, 2014

Re: [Pharo-dev] Roadmap on tools?

by Serge Stinckwich

On Mon, Aug 25, 2014 at 11:17 PM, Alain Rastoul <alr.dev(a)free.fr> wrote: > +1 for the GTInspector, and the Moose tools/paradigm too (Roassal, Glamour, > Moose and others), all that stuff is great step forward and could arouse > interest from doubtful people. > I remember myself failing to show some collegues at work how the smalltalk > system could be a cool tool to play with, even for people sticking on > dotNet, Delphi or C++. > And sometimes they remember that too ... > (Smalltalk? Squeak? -at that time- that blinking and poping toy ? hahaha > ...) > :( > Still working on that like a flea (?- a morpion) > > Morphic removed is good news - clumsy, buggy and weird - but I don't > understand the relationship with GTInspector ? > I googled about that and just found a post of you about Bloc in the mailing > list, it sounds like a good idea, and I'm sure you'll manage to do it > cleanly, but I'm also very curious about that: big bang or dependency > injection and small steps? other patterns, techniques ? a link on Bloc ? > I'm also curious about Spec and it's status after it's change to GPL ? Will > it be supported in the future ? What are the alternatives ? Yes, apparently spec is distributed now under a dual licence : MIT when used as an external library (not sure what it means when you use Smalltalk) and GPL when integrated in an IDE ... I think that this is a potential problem for Pharo. https://github.com/spec-framework/spec#license -- Serge Stinckwich UCBN & UMI UMMISCO 209 (IRD/UPMC) Every DSL ends up being Smalltalk http://www.doesnotunderstand.org/

Aug. 26, 2014

Re: [Pharo-dev] [Moose-dev] [ANN] Voronyj Diagram

by Natalia Tymchuk

Hello. Yes, it is quite slow. I used Delaunay triangulation to build the diagram, Iâm not sure that it has the smallest complexity, but it is the simples one for me. > I've seen > the code copies collections a lot, that might be itâ¦ It is the reason, however I wanted not only draw the diagram as a lot of segments, but to build the polygons and for that I need all that collections. Maybe I will change the algorithm, that builds the diagrams. Thanks for your attention and comments) Best regards, Natalia On Aug 26, 2014, at 3:38 AM, Damien Pollet <damien.pollet(a)gmail.com> wrote: > Hi Natalia, > > I've been playing with it today, and look at the pretty pixels it makes :D > The points are generated by the poisson disc sampling algorithm > described at http://bost.ocks.org/mike/algorithms/ > > However, see the result of the doit, in redâ¦ the time grows really > quick with the number of nodes, which makes it impractical with large > point sets (the 2 smaller sets are instant, the 3rd one quickly pops > the progress bar, and the 4th takes a couple minutes). Wikipedia says > the complexity should be around N log N, so that's strange ; I've seen > the code copies collections a lot, that might be itâ¦ but it's too late > for coding now :| > > On 25 July 2014 19:12, Natalia Tymchuk <natalia.tymchuk(a)unikernel.net> wrote: >> Thank you. >> Yes, I will. >> >> Natalia >> >> On Jul 25, 2014, at 5:36 PM, Alexandre Bergel <alexandre.bergel(a)me.com> >> wrote: >> >> Impressive!!!! >> >> Natalia, will you be at esug? >> >> Alexandre >> -- >> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >> Alexandre Bergel http://www.bergel.eu >> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >> >> >> >> On Jul 25, 2014, at 8:10 AM, Natalia Tymchuk <natalia.tymchuk(a)unikernel.net> >> wrote: >> >> Hello. >> >> I implemented Voronyj diagrams for Roassal. You can read about it here: >> http://natalia.tymchuk.me/RTVoronyjDiagram/ . >> I will be on vacation from Monday till Thursday, but it will be really cool >> if you can write me your opinion about this project and all your remarks. >> >> Best regards, >> Natalia >> _______________________________________________ >> Moose-dev mailing list >> Moose-dev(a)iam.unibe.ch >> https://www.iam.unibe.ch/mailman/listinfo/moose-dev >> >> >> > > > > -- > Damien Pollet > type less, do more [ | ] http://people.untyped.org/damien.pollet > <Poisson + Voronyj.png>_______________________________________________ > Moose-dev mailing list > Moose-dev(a)iam.unibe.ch > https://www.iam.unibe.ch/mailman/listinfo/moose-dev

Aug. 26, 2014

Re: [Pharo-dev] Roadmap on tools?

by phil＠highoctane.be

> > I strongly disagree on Morphic, personally I love using it. It can be > cleaner but still beats any other GUI I have used in terms of ease of use > without sacrificing power. Spec on the other hand looks ugly to me , weird > and very difficult to understand. But maybe its just my opinion and how my > brain is wired. Afterall Spec has been very popular with Pharo devs so > definitely is something I keep a close eye on . I really hope Morphic does > not get removed, at least continue to be distributed as third party > library.Though thats is unlikely with all the hard work of cleaning it up. > Morphic and Spec address different concerns. Building a UI with Morphic alone is what one would use to do something very custom (like a game for example). Now, creating a larger UI that way is definitely going to be super pain in the assets. That's where Spec does fit. Of course at this point in time, the underlying Morphic widgets are quite complicated and sometimes behave in strange ways (try right clicks on the dev tools in unexpected areas...). It takes a while to get used to but it works well. http://spec.st/docs/home/ I also wouldn't want Morphic to be removed, but definitely would like to see it cleaned up in some areas. What I am not sure about is how much of an external library we will depend on. One key thing in Morphic is that one can learn a lot by looking inside (and solve problems), something which will maybe not be there with a new version. Phil > > > On Tue, Aug 26, 2014 at 12:17 AM, Alain Rastoul <alr.dev(a)free.fr> wrote: > >> +1 for the GTInspector, and the Moose tools/paradigm too (Roassal, >> Glamour, Moose and others), all that stuff is great step forward and could >> arouse interest from doubtful people. >> I remember myself failing to show some collegues at work how the >> smalltalk system could be a cool tool to play with, even for people >> sticking on dotNet, Delphi or C++. >> And sometimes they remember that too ... >> (Smalltalk? Squeak? -at that time- that blinking and poping toy ? hahaha >> ...) >> :( >> Still working on that like a flea (?- a morpion) >> >> Morphic removed is good news - clumsy, buggy and weird - but I don't >> understand the relationship with GTInspector ? >> I googled about that and just found a post of you about Bloc in the >> mailing list, it sounds like a good idea, and I'm sure you'll manage to do >> it cleanly, but I'm also very curious about that: big bang or dependency >> injection and small steps? other patterns, techniques ? a link on Bloc ? >> I'm also curious about Spec and it's status after it's change to GPL ? >> Will it be supported in the future ? What are the alternatives ? >> >> I understand that you are very busy, so no problem if there is no link on >> existing material, I should survive, >> just wait a little :) >> >> TIA >> >> >> Regards >> >> Alain >> >> Le 24/08/2014 08:03, stepharo a Ã©crit : >> >> >>> On 24/8/14 05:39, Alexandre Bergel wrote: >>> >>>> I see GTInspector as a big splash. I guess that if there would be a >>>> Tool Roadmap, it would be focused on GTInspector. >>>> >>> >>> Not necessarily. >>> No time to explain sorry but else I would not improve morphic because >>> bloc is coming and parallel actions are important. >>> >>> Stef >>> >> >> >> >> >> >

Aug. 26, 2014

Re: [Pharo-dev] Videos ESUG2014

by Marcus Denker

Thanks, I will add an extra playlist for IWST On 25 Aug 2014, at 22:19, Sebastian Tleye <stleye(a)gmail.com> wrote: > Mmm, strange, here are the links (I don't know if there are any missing) > > Understanding Pharo's global state to move program through time and space > > Live robot programming > > Toward a methodology to turm Smalltalk code to FPGA > > Towards agile cross-platform development > > The Moldable Inspector > > ClockSystem: Embedding Time in Smalltalk > > Reducing Waste in Expandable Collections: The Pharo Case > > > https://www.youtube.com/watch?v=vJOvU1V5hZ4&list=UUO-vBhaKVZf0al-ISMMPvRw > > https://www.youtube.com/watch?v=hwUOwpidp0A&list=UUO-vBhaKVZf0al-ISMMPvRw > > https://www.youtube.com/watch?v=XWHq4IQ7gIQ&list=UUO-vBhaKVZf0al-ISMMPvRw > > https://www.youtube.com/watch?v=OkDOiGZ8R1E&list=UUO-vBhaKVZf0al-ISMMPvRw > > https://www.youtube.com/watch?v=V4CVp77yhUo&list=UUO-vBhaKVZf0al-ISMMPvRw > > https://www.youtube.com/watch?v=AUZQemqQphg&list=UUO-vBhaKVZf0al-ISMMPvRw > > > > > 2014-08-25 22:13 GMT+02:00 Chris Cunningham <cunningham.cb(a)gmail.com>: > Naive response, but no, I don't see them on YouTube. > Nor at > http://www.esug.org/wiki/pier/Conferences/2014 > > > On Mon, Aug 25, 2014 at 1:06 PM, Sebastian Tleye <stleye(a)gmail.com> wrote: > > > > 2014-08-25 1:39 GMT+02:00 Javier PimÃ¡s <elpochodelagente(a)gmail.com>: > > Are the papers also available? > > Yeap, You don't see them? > > Cheers, > Javier > > > On Sat, Aug 23, 2014 at 8:03 AM, stepharo <stepharo(a)free.fr> wrote: > this is why you should put info in the slides :) > > > On 23/8/14 00:02, Igor Stasenko wrote: > Haha.. > my name (and voice) is now [written] recorded [in stone] on internet. > One step closer to world domination! :) > > Too bad, i was so dragged by presenting the stuff, that i forgot to give a credit to the guy, sitting next to me, who put a lot of effort to make Virtual CPU work. SHAME ON ME! :/ > His name is Max Mattone - remember his name! > And the list won't be complete without Clement Bera, who also contributed a lot to it as well. And many others in RMoD team. > > -- > Best regards, > Igor Stasenko. > > > > > > -- > Javier PimÃ¡s > Ciudad de Buenos Aires > > >

Aug. 26, 2014

Re: [Pharo-dev] Videos ESUG2014

by Marcus Denker

On 25 Aug 2014, at 22:22, Sebastian Tleye <stleye(a)gmail.com> wrote: > Now, if you meant the paper but not the presentation, then I don't know where they are :-) > -> I will request all slides that are missing -> I will sync with the Student Volunteers for Videos -> Papers will be available, too. (I will check with the workshop organisers) > > 2014-08-25 22:19 GMT+02:00 Sebastian Tleye <stleye(a)gmail.com>: > Mmm, strange, here are the links (I don't know if there are any missing) > > Understanding Pharo's global state to move program through time and space > > Live robot programming > > Toward a methodology to turm Smalltalk code to FPGA > > Towards agile cross-platform development > > The Moldable Inspector > > ClockSystem: Embedding Time in Smalltalk > > Reducing Waste in Expandable Collections: The Pharo Case > > > https://www.youtube.com/watch?v=vJOvU1V5hZ4&list=UUO-vBhaKVZf0al-ISMMPvRw > > https://www.youtube.com/watch?v=hwUOwpidp0A&list=UUO-vBhaKVZf0al-ISMMPvRw > > https://www.youtube.com/watch?v=XWHq4IQ7gIQ&list=UUO-vBhaKVZf0al-ISMMPvRw > > https://www.youtube.com/watch?v=OkDOiGZ8R1E&list=UUO-vBhaKVZf0al-ISMMPvRw > > https://www.youtube.com/watch?v=V4CVp77yhUo&list=UUO-vBhaKVZf0al-ISMMPvRw > > https://www.youtube.com/watch?v=AUZQemqQphg&list=UUO-vBhaKVZf0al-ISMMPvRw > > > > > 2014-08-25 22:13 GMT+02:00 Chris Cunningham <cunningham.cb(a)gmail.com>: > > Naive response, but no, I don't see them on YouTube. > Nor at > http://www.esug.org/wiki/pier/Conferences/2014 > > > On Mon, Aug 25, 2014 at 1:06 PM, Sebastian Tleye <stleye(a)gmail.com> wrote: > > > > 2014-08-25 1:39 GMT+02:00 Javier PimÃ¡s <elpochodelagente(a)gmail.com>: > > Are the papers also available? > > Yeap, You don't see them? > > Cheers, > Javier > > > On Sat, Aug 23, 2014 at 8:03 AM, stepharo <stepharo(a)free.fr> wrote: > this is why you should put info in the slides :) > > > On 23/8/14 00:02, Igor Stasenko wrote: > Haha.. > my name (and voice) is now [written] recorded [in stone] on internet. > One step closer to world domination! :) > > Too bad, i was so dragged by presenting the stuff, that i forgot to give a credit to the guy, sitting next to me, who put a lot of effort to make Virtual CPU work. SHAME ON ME! :/ > His name is Max Mattone - remember his name! > And the list won't be complete without Clement Bera, who also contributed a lot to it as well. And many others in RMoD team. > > -- > Best regards, > Igor Stasenko. > > > > > > -- > Javier PimÃ¡s > Ciudad de Buenos Aires > > > >

Aug. 26, 2014

Re: [Pharo-dev] Failing make a slice

by Marcus Denker

On 26 Aug 2014, at 00:18, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > A slice is nothing more than an empty mcz package with as requirements some other packages, namely the ones that you changed and want to include. I guess that if you take a fresh image and merge in your packages, you can create a new slice over them (provided they are dirty, sometimes this is a bit hard, you could make a trivial change). Make sure to save the slice with a higher version number. > The problem that the bug seems deeperâ¦ I have the impression that âSave Sliceâ does not search all the repos and if the number already exists, increase (like normal slice saving). Then, when saving, it *saves* the wrong version as it asks the cache âgive me name Xâ, and âname Xâ is already there, and it uses that for upload (or even skips the upload as the file is already there?). Even re-saving the slice does not help, as unchanged members are not re-saved with a new name. But even if this would work, everything fails as soon as you end up with two MCZ files with the same name (e.g. working in two images offline, or working on a package, save, but then only merge it *after* you did another change in the main branch that led to the same name). Maybe we should really add the hash in the filenameâ¦ Marcus > HTH > > On 25 Aug 2014, at 22:45, Juraj Kubelka <juraj.kubelka(a)gmail.com> wrote: > >> Is there a way to load the right packages and create a new slice? >> >> Thanks, >> Juraj >> >> On Aug 25, 2014, at 3:21 PM, stepharo <stepharo(a)free.fr> wrote: >> >>> I'm really thinking that we should use a UUID in addition to author names to avoid such kind of problem. >>>> Hi, >>>> >>>> I have failed to make new slice for this issue: https://pharo.fogbugz.com/f/cases/13185/Move-embedded-fonts-from-Freetype-p… >>>> >>>> The last slice I have made depends on packages which has the same name but different UUID. >>>> >>>> How can I fix this? The issue should depend on FreeType-JurajKubelka.677, EmbeddedFreeType-JurajKubelka.3, EmbeddedFreeTypeTests-JurajKubelka.2. >>>> >>>> Thanks, >>>> Juraj >>>> >>> >>> >> >> > >

Aug. 26, 2014

Re: [Pharo-dev] [Moose-dev] [ANN] Voronyj Diagram

by Damien Pollet

Hi Natalia, I've been playing with it today, and look at the pretty pixels it makes :D The points are generated by the poisson disc sampling algorithm described at http://bost.ocks.org/mike/algorithms/ However, see the result of the doit, in redâ¦ the time grows really quick with the number of nodes, which makes it impractical with large point sets (the 2 smaller sets are instant, the 3rd one quickly pops the progress bar, and the 4th takes a couple minutes). Wikipedia says the complexity should be around N log N, so that's strange ; I've seen the code copies collections a lot, that might be itâ¦ but it's too late for coding now :| On 25 July 2014 19:12, Natalia Tymchuk <natalia.tymchuk(a)unikernel.net> wrote: > Thank you. > Yes, I will. > > Natalia > > On Jul 25, 2014, at 5:36 PM, Alexandre Bergel <alexandre.bergel(a)me.com> > wrote: > > Impressive!!!! > > Natalia, will you be at esug? > > Alexandre > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > On Jul 25, 2014, at 8:10 AM, Natalia Tymchuk <natalia.tymchuk(a)unikernel.net> > wrote: > > Hello. > > I implemented Voronyj diagrams for Roassal. You can read about it here: > http://natalia.tymchuk.me/RTVoronyjDiagram/ . > I will be on vacation from Monday till Thursday, but it will be really cool > if you can write me your opinion about this project and all your remarks. > > Best regards, > Natalia > _______________________________________________ > Moose-dev mailing list > Moose-dev(a)iam.unibe.ch > https://www.iam.unibe.ch/mailman/listinfo/moose-dev > > > -- Damien Pollet type less, do more [ | ] http://people.untyped.org/damien.pollet

Aug. 26, 2014

athens wrapper for (Form-)Canvas

by Nicolai Hess

I tried an alternative approach for rendering Morphs on athens without implementing drawOnAthensCanvas for every morph: 13866 <https://pharo.fogbugz.com/default.asp?13866> AthensFormCanvasWrapper still struggling with the text rendering. Yet I am not sure what is the better approach. Nicolai

Aug. 25, 2014

Re: [Pharo-dev] Failing make a slice

by Sven Van Caekenberghe

A slice is nothing more than an empty mcz package with as requirements some other packages, namely the ones that you changed and want to include. I guess that if you take a fresh image and merge in your packages, you can create a new slice over them (provided they are dirty, sometimes this is a bit hard, you could make a trivial change). Make sure to save the slice with a higher version number. HTH On 25 Aug 2014, at 22:45, Juraj Kubelka <juraj.kubelka(a)gmail.com> wrote: > Is there a way to load the right packages and create a new slice? > > Thanks, > Juraj > > On Aug 25, 2014, at 3:21 PM, stepharo <stepharo(a)free.fr> wrote: > >> I'm really thinking that we should use a UUID in addition to author names to avoid such kind of problem. >>> Hi, >>> >>> I have failed to make new slice for this issue: https://pharo.fogbugz.com/f/cases/13185/Move-embedded-fonts-from-Freetype-p… >>> >>> The last slice I have made depends on packages which has the same name but different UUID. >>> >>> How can I fix this? The issue should depend on FreeType-JurajKubelka.677, EmbeddedFreeType-JurajKubelka.3, EmbeddedFreeTypeTests-JurajKubelka.2. >>> >>> Thanks, >>> Juraj >>> >> >> > >

Aug. 25, 2014