May 2011 - Pharo-dev - lists.pharo.org

Re: [Pharo-project] Zinc feedback

by Igor Stasenko

On 20 May 2011 10:08, Sven Van Caekenberghe <sven(a)beta9.be> wrote: > Stef, > > On 20 May 2011, at 08:57, StÃ©phane Ducasse wrote: > >> Hi Sven >> >> with igor we went over the start up list and we saw that ZnServer resgiter all subclasses of ZnServer. >> We think that it would be better to have a list of servers to be registered and that users can decide. >> What do you think? >> >> Stef > > It is the traditional/classic idea that a server object in Smalltalk should survive image save/load. This is very convenient for beginners and during development. For production systems, explicit control is better. > > Now that I look at the code again, I must agree that it is a bit weird: only those server registered/started as the default in their class fall under this scheme, not other instances. > > So yes, an explicit managed list of ZnServer subclasses that need startup/shutdown would be nicer. If that is what you mean. > Yes. This is what i meant :) Because Default means only single server. While actually there could be as many servers as you want. So, during image startup, there should be a registration mechanism, which registers servers and starting them up. Instead of 'Default' single instance. > Sven > -- Best regards, Igor Stasenko AKA sig.

May 20, 2011

Re: [Pharo-project] bug in Nautilus/RPackage?

by Cyrille Delaunay

I don't know how to reproduce the situation in the screenshot, but I have a small idea of what the problem is: Your package Spy is composed of several subcategories. All the classes are defined in the subcategories but none is defined in the one called 'Spy'. In Rpackage, the classes are put inside the RPackage matching the category in which the class is defined 2011/5/19 Alexandre Bergel <alexandre.bergel(a)me.com> > Thanks for making it happen :-) > > Alexandre > > > On 19 May 2011, at 15:39, StÃ©phane Ducasse wrote: > > > thanks for reporting it. > > > > do you have a reproducable way of arriving to this state? > > > >> Hi ! > >> > >> I think I bumped into a bug when using Nautilus. I am not sure whether > this is due to RPackage or not. > >> > >> I loaded Spy, and apparently there is no class in the package within > Nautilus. > >> > >> Cheers, > >> Alexandre > >> -- > >> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > >> Alexandre Bergel http://www.bergel.eu > >> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > >> > >> > >> > >> > >> <Screen shot 2011-05-19 at 11.21.23.png> > > > > > > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > > > > >

May 20, 2011

Re: [Pharo-project] Zinc feedback

by Sven Van Caekenberghe

Stef, On 20 May 2011, at 08:57, StÃ©phane Ducasse wrote: > Hi Sven > > with igor we went over the start up list and we saw that ZnServer resgiter all subclasses of ZnServer. > We think that it would be better to have a list of servers to be registered and that users can decide. > What do you think? > > Stef It is the traditional/classic idea that a server object in Smalltalk should survive image save/load. This is very convenient for beginners and during development. For production systems, explicit control is better. Now that I look at the code again, I must agree that it is a bit weird: only those server registered/started as the default in their class fall under this scheme, not other instances. So yes, an explicit managed list of ZnServer subclasses that need startup/shutdown would be nicer. If that is what you mean. Sven

May 20, 2011

[Pharo-project] [update 1.3] #13205

by Marcus Denker

13205 ----- Issue 4249: ProcessBrowser Browsers classvar is dormant Issue 4247: Remove MenuIcons from startUpList Issue 4250: ImageSegment - remove from startup list Issue 4251: DateAndTime - remove from startup list Issue 4246: Avoid harcoding menu in MCWorkingCopyMenu -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.

May 20, 2011

Re: [Pharo-project] Set with nil

by StÃ©phane Ducasse

On May 20, 2011, at 9:40 AM, Bernat Romagosa wrote: > If we understand nil as the empty set (like some other languages do), ??? No nil is not the empty set this is the undefined object. Stef > then Set with: nil is not any different than Set with: Set new. > > 2011/5/20 Igor Stasenko <siguctua(a)gmail.com> > On 20 May 2011 04:13, Guillermo Polito <guillermopolito(a)gmail.com> wrote: > > Hi! > > > > Since a year ago, we have this issue: > > > > http://code.google.com/p/pharo/issues/detail?id=1907 > > > > There were 2 solutions proposed, and the 2 were developed by different > > people. But for some reason it was not integrated and continuously > > postponed. > > > > So, my question is: Do we want Sets accepting nil? If we don't, the issue > > is invalid. If we do, can someone review the code proposed? > > > > Its not invalid. Stef reminds me about it from time to time. > And actually we wanted to sit and review and integrate it but it > always something else needs our attention. > > > Thanks! > > Guille > > > > > > -- > Best regards, > Igor Stasenko AKA sig. > >

May 20, 2011

Re: [Pharo-project] Name spaces

by StÃ©phane Ducasse

On May 20, 2011, at 9:40 AM, Noury Bouraqadi wrote: > Hi, > > I'm wondering what is the status of the name spaces project developed last summer as part of GSoC. nowhere and do not expect it to be in pharo. Stef > > Noury > http://car.mines-douai.fr/noury > -- > -6th National Conference on > âControl Architecture of Robotsâ > 24-25 may 2011, Grenoble area, France > http://car2011.inrialpes.fr/ > > -19th ESUG International Smalltalk Conference > 22-26 August 2011, Edinburgh, UK > http://www.esug.org/Conferences/2011 > > -19Ã¨mes JournÃ©es Francophones sur les SystÃ¨mes Multi-Agents (JFSMAâ11) > http://www.univ-valenciennes.fr/congres/jfsma2011/ > 17-19 Octobre 2011, Valenciennes, France > > > > > >

May 20, 2011

Re: [Pharo-project] Set with nil

by Bernat Romagosa

If we understand nil as the empty set (like some other languages do), then Set with: nil is not any different than Set with: Set new. 2011/5/20 Igor Stasenko <siguctua(a)gmail.com> > On 20 May 2011 04:13, Guillermo Polito <guillermopolito(a)gmail.com> wrote: > > Hi! > > > > Since a year ago, we have this issue: > > > > http://code.google.com/p/pharo/issues/detail?id=1907 > > > > There were 2 solutions proposed, and the 2 were developed by different > > people. But for some reason it was not integrated and continuously > > postponed. > > > > So, my question is: Do we want Sets accepting nil? If we don't, the > issue > > is invalid. If we do, can someone review the code proposed? > > > > Its not invalid. Stef reminds me about it from time to time. > And actually we wanted to sit and review and integrate it but it > always something else needs our attention. > > > Thanks! > > Guille > > > > > > -- > Best regards, > Igor Stasenko AKA sig. > >

May 20, 2011

[Pharo-project] Name spaces

by Noury Bouraqadi

Hi, I'm wondering what is the status of the name spaces project developed last summer as part of GSoC. Noury http://car.mines-douai.fr/noury -- -6th National Conference on âControl Architecture of Robotsâ 24-25 may 2011, Grenoble area, France http://car2011.inrialpes.fr/ -19th ESUG International Smalltalk Conference 22-26 August 2011, Edinburgh, UK http://www.esug.org/Conferences/2011 -19Ã¨mes JournÃ©es Francophones sur les SystÃ¨mes Multi-Agents (JFSMAâ11) http://www.univ-valenciennes.fr/congres/jfsma2011/ 17-19 Octobre 2011, Valenciennes, France

May 20, 2011

Re: [Pharo-project] problem with Dictionary and KeyNowFound

by StÃ©phane Ducasse

I would like to have tests covering that! On May 20, 2011, at 12:34 AM, Camillo Bruni wrote: > > On 2011-05-20, at 00:03, Mariano Martinez Peck wrote: > >> On Thu, May 19, 2011 at 11:55 PM, Camillo Bruni <camillo.bruni(a)inria.fr>wrote: >> >>> IMO this shouldn't change a thing, since KeyNotFound inherits from Error... >>> having a look at it >>> >>> yes i know :( >> but in pharo 1.2 it works. That's why I thought it could be related to >> that. > > ok, the bug seems to be related to catching the signaller, dunno why, but commenting out the section on Exception >> signal seems to solve the issue. > > I will have a look at it tomorrow. > > camillo > >>> On 2011-05-19, at 23:45, Mariano Martinez Peck wrote: >>>> So...it seems Pharo1.3 change Dictionary to throw KeyNotFound instead of >>> Error. >>>> Now, my image freezes with a loop. I would appreciate some help. I >>> couldn't really analyzed the problem. >>>> Fileine both FLPerson.st and ATest and run the test testDictionaryRehash. >>>> In such test, a the end there is a >>>> >>>> self shouldnt: [aDictionary at: newPerson1] raise: KeyNotFound. >>>> >>>> >>>> which raises the debugger pre-popup. if I try to click on the popup... -> >>> loop/freeze. >>>> You can interrupt it and see more or less what is happening. >>>> >>>> Camillo, can you take a look? >>> >> >> >> >> -- >> Mariano >> http://marianopeck.wordpress.com > >

May 20, 2011

Re: [Pharo-project] Set with nil

by Igor Stasenko

On 20 May 2011 04:13, Guillermo Polito <guillermopolito(a)gmail.com> wrote: > Hi! > > Since a year ago, we have this issue: > > http://code.google.com/p/pharo/issues/detail?id=1907 > > There were 2 solutions proposed, and the 2 were developed by different > people.Â But for some reason it was not integrated and continuously > postponed. > > So, my question is: Do we want Sets accepting nil?Â If we don't, the issue > is invalid.Â If we do, can someone review the code proposed? > Its not invalid. Stef reminds me about it from time to time. And actually we wanted to sit and review and integrate it but it always something else needs our attention. > Thanks! > Guille > -- Best regards, Igor Stasenko AKA sig.

May 20, 2011