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

Re: [Pharo-dev] Smalltalk Impact

by Esteban A. Maringolo

2014/1/17 Sven Van Caekenberghe <sven(a)stfx.eu> > > >> This is usual attitude in corporate environment: you paying me to do the > >> job, > >> i do it and i don't care about the rest. Been there, ate that :) > > > > If I pay you to do X, and you spend all my money doing Y, you need a > > very good story explaining why I don't get what I paid for. > > If I get paid to do X, based on an open-source platform that helps me tremendously, then it is only logical and ethical to give back and support that platform - it would even be in my own self interest. This holds true as long as you deliver X. Some companies to delivery X build Y and Z, and open source Y & Z [1] Esteban A. Maringolo [1] E.g. http://square.github.io/

Jan. 17, 2014

Re: [Pharo-dev] Smalltalk Impact

by Frank Shearar

On 17 January 2014 10:49, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > > On 17 Jan 2014, at 11:27, Frank Shearar <frank.shearar(a)gmail.com> wrote: > >> On 16 January 2014 23:16, Igor Stasenko <siguctua(a)gmail.com> wrote: >>> >>> >>> >>> On 16 January 2014 21:15, Hilaire Fernandes <hilaire.fernandes(a)gmail.com> >>> wrote: >>>> >>>> Le 16/01/2014 12:55, H. Hirzel a Ã©crit : >>>> >>>>>>> âWe are not payed to work on Squeakâ, that is what I got toldâ¦ >>>>> >>>>> Probably correct at that time... >>>> >>> This is usual attitude in corporate environment: you paying me to do the >>> job, >>> i do it and i don't care about the rest. Been there, ate that :) >> >> If I pay you to do X, and you spend all my money doing Y, you need a >> very good story explaining why I don't get what I paid for. > > If I get paid to do X, based on an open-source platform that helps me tremendously, then it is only logical and ethical to give back and support that platform - it would even be in my own self interest. > > Big, hot US companies do this all the time for thousands of projects. That is true of the entity paying to do the work. If you paid me to get features X, Y & Z done, and instead I did lots of (good!) work on other projects _not directly relevant to X, Y & Z_, I think you'd have every right to get angry with me. I'm _not_ suggesting that it's right for someone to completely ignore the libraries that they use. I am saying that there are multiple forces behind "we don't get paid to do X", and "this work doesn't help me deliver X" is perfectly legitimate. frank >> Seen in hindsight, everything is obvious. At the time it might have >> been far from obvious. >> >> frank > >

Jan. 17, 2014

Re: [Pharo-dev] IncompatibleLayoutConflict with NativeBoost in recent #30700

by Camille Teruel

On 17 janv. 2014, at 11:13, Igor Stasenko <siguctua(a)gmail.com> wrote: > > > > On 17 January 2014 10:09, Camille Teruel <camille.teruel(a)gmail.com> wrote: > > On 16 janv. 2014, at 16:58, Marcus Denker <marcus.denker(a)inria.fr> wrote: > >> Hello, >> >> I was looking into that, the first (maybe only) reason is that the new class builder >> can not make anonymous subclasses of classes with ByteLayout > > Is that really a bug? > Should it really be possible to subclass a class that has a byte layout with a class that has a fixed layout? > > > i don't know and i don't care. > Why i should care about such details in NB while all i want is to create simple anonymous subclass? > > > Because: > AnonymousClassInstaller make: [ :builder | > builder superclass: NBExternalTypeValue ] > raises IncompatibleLayout but: > AnonymousClassInstaller make: [ :builder | > builder > superclass: NBExternalTypeValue; > layoutClass: ByteLayout ] > doesn't. > A simple fix is to take the same layout class than the superclass by default instead of FixedLayout, isn't it? > For example with lazy initialization: > > SlotClassBuilder>>#layoutClass > ^ layoutClass ifNil: [ layoutClass := superclass layout class ] > > > i don't understand what is difference between 'fixed' and 'same as superclass' layout, > and i am not sure i want to understand it. I want to be freed from knowing such details when doing something as simple as creating anonymous subclass. > > P.S. If you think it will work, feel free to post it. It works and it's commited. Waiting for the monkey and reviews: https://pharo.fogbugz.com/f/cases/12488/Can-not-create-newAnonymousSubclass… > > > -- > Best regards, > Igor Stasenko.

Jan. 17, 2014

Re: [Pharo-dev] Smalltalk Impact

by Sven Van Caekenberghe

On 17 Jan 2014, at 11:27, Frank Shearar <frank.shearar(a)gmail.com> wrote: > On 16 January 2014 23:16, Igor Stasenko <siguctua(a)gmail.com> wrote: >> >> >> >> On 16 January 2014 21:15, Hilaire Fernandes <hilaire.fernandes(a)gmail.com> >> wrote: >>> >>> Le 16/01/2014 12:55, H. Hirzel a Ã©crit : >>> >>>>>> âWe are not payed to work on Squeakâ, that is what I got toldâ¦ >>>> >>>> Probably correct at that time... >>> >> This is usual attitude in corporate environment: you paying me to do the >> job, >> i do it and i don't care about the rest. Been there, ate that :) > > If I pay you to do X, and you spend all my money doing Y, you need a > very good story explaining why I don't get what I paid for. If I get paid to do X, based on an open-source platform that helps me tremendously, then it is only logical and ethical to give back and support that platform - it would even be in my own self interest. Big, hot US companies do this all the time for thousands of projects. > Seen in hindsight, everything is obvious. At the time it might have > been far from obvious. > > frank

Jan. 17, 2014

ConfigurationOfGetText broken?

by Usman Bhatti

Hello, It seems that some recent changes in the ConfigurationOfText have broken the loading in the Pharo 3.0. I'm loading GetText with the following script in my config. spec project: 'GetText' with: [ spec className: 'ConfigurationOfGettext'; file: 'ConfigurationOfGettext'; loads: #('Core' ); version: '1.3'; "tried #development too" repository: 'http://smalltalkhub.com/mc/PharoExtras/Gettext/main']. I get this error whatever version I put. 'The version ''1.3'' is not defined in ConfigurationOfGettext for the current platform, because an exception occurred while creating the version:. Error: The imported version:''1.1-baseline'' for version: 1.3 referenced from the method: #version13: in configuration ConfigurationOfGettext has not been defined.. Evaluate the following to see the error: ''[ConfigurationOfGettext project ] on: MetacelloErrorInProjectConstructionNotification do: [:ex | ex resume: true ].'' Possible versions include: #(#bleedingEdge #development #stable ''1.0-baseline'' ''1.0'' ''1.1'' ''1.2'')' I tried to debug and compared with the changes from the previous version but didn't find any workable solution. Usman

Jan. 17, 2014

Re: [Pharo-dev] Smalltalk Impact

by Frank Shearar

On 16 January 2014 23:16, Igor Stasenko <siguctua(a)gmail.com> wrote: > > > > On 16 January 2014 21:15, Hilaire Fernandes <hilaire.fernandes(a)gmail.com> > wrote: >> >> Le 16/01/2014 12:55, H. Hirzel a Ã©crit : >> >> >>> âWe are not payed to work on Squeakâ, that is what I got toldâ¦ >> > >> > Probably correct at that time... >> > This is usual attitude in corporate environment: you paying me to do the > job, > i do it and i don't care about the rest. Been there, ate that :) If I pay you to do X, and you spend all my money doing Y, you need a very good story explaining why I don't get what I paid for. Seen in hindsight, everything is obvious. At the time it might have been far from obvious. frank

Jan. 17, 2014

Re: [Pharo-dev] IncompatibleLayoutConflict with NativeBoost in recent #30700

by Marcus Denker

On 17 Jan 2014, at 11:13, Igor Stasenko <siguctua(a)gmail.com> wrote: > > > > On 17 January 2014 10:09, Camille Teruel <camille.teruel(a)gmail.com> wrote: > > On 16 janv. 2014, at 16:58, Marcus Denker <marcus.denker(a)inria.fr> wrote: > >> Hello, >> >> I was looking into that, the first (maybe only) reason is that the new class builder >> can not make anonymous subclasses of classes with ByteLayout > > Is that really a bug? > Should it really be possible to subclass a class that has a byte layout with a class that has a fixed layout? > > > i don't know and i don't care. > Why i should care about such details in NB while all i want is to create simple anonymous subclass? > > > Because: > AnonymousClassInstaller make: [ :builder | > builder superclass: NBExternalTypeValue ] > raises IncompatibleLayout but: > AnonymousClassInstaller make: [ :builder | > builder > superclass: NBExternalTypeValue; > layoutClass: ByteLayout ] > doesn't. > A simple fix is to take the same layout class than the superclass by default instead of FixedLayout, isn't it? > For example with lazy initialization: > > SlotClassBuilder>>#layoutClass > ^ layoutClass ifNil: [ layoutClass := superclass layout class ] > > > i don't understand what is difference between 'fixed' and 'same as superclass' layout, > and i am not sure i want to understand it. I want to be freed from knowing such details when doing something as simple as creating anonymous subclass. This is a bugâ¦ But the fun thing is: I always explain to people that adding a second better subsystem but then not using it everywhere is not good. And we found another case! A bug in the new class builder that we would have found 5 months ago had we used the new one and not the old one in this case. Marcus

Jan. 17, 2014

Re: [Pharo-dev] IncompatibleLayoutConflict with NativeBoost in recent #30700

by Igor Stasenko

On 17 January 2014 10:09, Camille Teruel <camille.teruel(a)gmail.com> wrote: > > On 16 janv. 2014, at 16:58, Marcus Denker <marcus.denker(a)inria.fr> wrote: > > Hello, > > I was looking into that, the first (maybe only) reason is that the new > class builder > can not make anonymous subclasses of classes with ByteLayout > > > Is that really a bug? > Should it really be possible to subclass a class that has a byte layout > with a class that has a fixed layout? > > i don't know and i don't care. Why i should care about such details in NB while all i want is to create simple anonymous subclass? > Because: > AnonymousClassInstaller make: [ :builder | > builder superclass: NBExternalTypeValue ] > raises IncompatibleLayout but: > AnonymousClassInstaller make: [ :builder | > builder > superclass: NBExternalTypeValue; > layoutClass: ByteLayout ] > doesn't. > A simple fix is to take the same layout class than the superclass by > default instead of FixedLayout, isn't it? > For example with lazy initialization: > > SlotClassBuilder>>#layoutClass > ^ layoutClass ifNil: [ layoutClass := superclass layout class ] > i don't understand what is difference between 'fixed' and 'same as superclass' layout, and i am not sure i want to understand it. I want to be freed from knowing such details when doing something as simple as creating anonymous subclass. P.S. If you think it will work, feel free to post it. -- Best regards, Igor Stasenko.

Jan. 17, 2014

Re: [Pharo-dev] STHub: 1190 repositories, 1001 users registered and 110365 packages uploaded

by Yuriy Tymchuk

Letâs start from http://youtu.be/z11U4jJ7we4?hd=1 And also, Gephi is a tool for complex networks/graph visualisation. Roassal should be a tool for agile visualisation and together with Moose can provide powerful reverse-engenearing support. Idâ really like some improved interaction API in future releases. Cheers. Uko On 17 Jan 2014, at 02:51, Alexandre Bergel <alexandre.bergel(a)me.com> wrote: > Truly beautiful. Youâre raising some challenges for Roassal2 and Roassal2@Amber > > Alexandre > > > On Jan 16, 2014, at 10:00 PM, Yuriy Tymchuk <yuriy.tymchuk(a)me.com> wrote: > >> Yes, have you seen it, http://yuriy.tymchuk.me/Smalldromeda/ ? >> >> On 17 Jan 2014, at 00:32, Torsten Bergmann <astares(a)gmx.de> wrote: >> >>> see http://smalltalkhub.com >>> >> >> > > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > >

Jan. 17, 2014

[pharo-project/pharo-core]

by GitHub

Branch: refs/tags/30702 Home: https://github.com/pharo-project/pharo-core

Jan. 17, 2014