April 2015 - Pharo-dev - lists.pharo.org

Re: [Pharo-dev] Bloc Roadmap

by Alain Plantec

> > >> >> > Weâve implemented a lot of examples. please have a look. >> Thanks - I have and they are also great. But AFAICT it's impossible to tell from the examples /why/ e.g. we went from Morphs owning their step actions to block-based steps registered with a scheduler that is universal to all worlds. I have a doubt about that because one use case I've had in Morphic is to speed up time in a world to see what happens e.g. in 10 minutes from now, but #schedule hard-codes the reference to the BlUniverse singleton. Morphic uses two global queues (hold by the world state) one for stepping and the other for the alarms. bloc have only one global queue for stepping and alarms. basically the same kind of implementation as in morph. and again, any class can declare and manage its own AlarmQueues. :) Alain

April 1, 2015

Got a strange font bug

by stepharo

HI we were improving the selection of instance variables of the refactorings via nautilus when suddenly the font got broken. We got a problem with strikefont holding a number of something like that. Impossible to debug - it would be great to have a rescue path (may be via the EmergencyEvalutor) - the only thing that we could do was to kill the image and restart I suspect a problem with the FTPlugin. Stef

April 1, 2015

Re: [Pharo-dev] [ANN] Pharo Consortium New Academic Partner: UMMISCO

by stepharo

Welcome UMMISCO Stef Le 1/4/15 14:47, Marcus Denker a Ã©crit : > The Pharo Consortium is very happy to announce that UMMISCO > has joined the Consortium as an Academic Partner. > > UMMISCO is an international research unit under the supervision of IRD > (Research for Development) and UPMC (UniversitÃ© Pierre Marie Curie) > located in France, Senegal, Morocco, Cameroun and Vietnam. UMMISCO > group carries out research in Computer Science and Mathematics with > the goal of modelling and simulation of complex systems with > applications regarding southern countries concerns like : epidemiology > modelling, disaster management, urban mobility and environmental > monitoring. > > About > > - UMMISCO: http://www.ummisco.ird.fr > - Pharo Consortium: http://consortium.pharo.org > > The goal of the Pharo Consortium is to allow companies and institutions to > support the ongoing development and future of Pharo. > Individuals can support Pharo via the Pharo Association: > > http://association.pharo.org >

April 1, 2015

Re: [Pharo-dev] Bloc Roadmap

by Alain Plantec

> On 01 Apr 2015, at 22:15, Sean P. DeNigris <sean(a)clipperadams.com> wrote: > > > yes, we know, we are preparing a documentation. > Great! Is the WIP publicly available (e.g. in one of our Pillar/GitHub repos)? Btw. the class comments are reeeally great! I'm only talking about documenting design rationale. https://github.com/SquareBracketAssociates/PharoInProgress/tree/master/Bloc <https://github.com/SquareBracketAssociates/PharoInProgress/tree/master/Bloc> see blocCore.pillar work in progress... > > > Weâve implemented a lot of examples. please have a look. > Thanks - I have and they are also great. But AFAICT it's impossible to tell from the examples /why/ e.g. we went from Morphs owning their step actions to block-based steps registered with a scheduler that is universal to all worlds. I have a doubt about that because one use case I've had in Morphic is to speed up time in a world to see what happens e.g. in 10 minutes from now, but #schedule hard-codes the reference to the BlUniverse singleton. Bloc use Alarms, Stepping will be removed. Each space has its own queue of alarms but nothing prevent a morph to have its own queue. Now, steps and alarms are not limited to use blocks, one can use message send also. Cheers Alain > Cheers, > Sean > > View this message in context: Re: Bloc Roadmap <http://forum.world.st/Bloc-Roadmap-tp4816171p4816769.html> > Sent from the Pharo Smalltalk Developers mailing list archive <http://forum.world.st/Pharo-Smalltalk-Developers-f1294837.html> at Nabble.com.

April 1, 2015

Re: [Pharo-dev] Bloc Roadmap

by stepharo

> > yes, we know, we are preparing a documentation. > Great! Is the WIP publicly available (e.g. in one of our Pillar/GitHub > repos)? Btw. the class comments are reeeally great! I'm only talking > about documenting design rationale. We did them twice :) And some of them should be improved. To doc in on github pharo progress. Now I started to document the classes so that after I can generate part of the doc :) > > Weâve implemented a lot of examples. please have a look. > Thanks - I have and they are also great. But AFAICT it's impossible to > tell from the examples /why/ e.g. we went from Morphs owning their > step actions to block-based steps registered with a scheduler that is > universal to all worlds. I have a doubt about that because one use > case I've had in Morphic is to speed up time in a world to see what > happens e.g. in 10 minutes from now, but #schedule hard-codes the > reference to the BlUniverse singleton. Good question alain will answer. Continue to ask so that we can extract the information from his mind. > Cheers, > Sean > > ------------------------------------------------------------------------ > View this message in context: Re: Bloc Roadmap > <http://forum.world.st/Bloc-Roadmap-tp4816171p4816769.html> > Sent from the Pharo Smalltalk Developers mailing list archive > <http://forum.world.st/Pharo-Smalltalk-Developers-f1294837.html> at > Nabble.com.

April 1, 2015

Re: [Pharo-dev] PharoV30.sources in Pharo 4.0 download

by stepharo

I think so. Stef Le 1/4/15 22:23, Chris Cunningham a Ã©crit : > Is there a plan to condense changes and build a new sources file for > the 4.0 release? > > Just curious > -cbc > > On Wed, Apr 1, 2015 at 1:18 PM, Max Leske <maxleske(a)gmail.com > <mailto:maxleske@gmail.com>> wrote: > > >> On 01 Apr 2015, at 22:12, Dmitri Zagidulin <dmitri(a)zagidulin.net >> <mailto:dmitri@zagidulin.net>> wrote: >> >> I noticed that the Pharo 4.0 full distribution zip file off of >> Pharo.org/download <http://Pharo.org/download> contains >> PharoV30.sources (and not, say, PharoV40.sources). >> This is true for all 3 platforms: >> - http://files.pharo.org/platform/Pharo4.0-mac.zip >> - http://files.pharo.org/platform/Pharo4.0-win.zip >> - http://files.pharo.org/platform/Pharo4.0-linux.zip >> >> Is there a typo in a CI job somewhere? > > No, those are the latest sources :) The version simply tells you > with what image the sources have been created (I presume). Anyway, > the version number is not important. > >

April 1, 2015

Re: [Pharo-dev] Enhancing the AST API

by stepharo

I do not know. personally shorter is not a good measure to me. When I spend time understanding code I prefer simpler to understand than cryptic. Now I will have to check what is stored exactly and its meaning. Stef > what about leftIndex? it's shorter > > On Tue, Mar 31, 2015 at 12:53 PM, Thierry Goubier > <thierry.goubier(a)gmail.com> wrote: >> Le 31/03/2015 11:04, stepharo a Ã©crit : >>> Hi >>> >>> in Pharo 50 I would like to improve the AST API. >>> often we do not know if an instance variable holds a position or a node >>> and this should be clear. >>> For example I propose to have leftPosition instead of left. >> >> If you like. >> >>> Does anybody has a better suggestion? >> >> Well, I'm allways under the assumption that an instance variable is internal >> only and that, when querying an AST node, I should allways use its API (is >> that usually #startPosition and #stopPosition?). >> >>> This is really annoying to have to look at the users of an instance >>> variable to understand its role. >> >> Or have a well commented class to understand its role? It is not allways >> possible to have instantaneously understandable instance variable names. >> >> Thierry >> >> > >

April 1, 2015

Re: [Pharo-dev] Enhancing the AST API

by stepharo

> Well, I'm allways under the assumption that an instance variable is > internal only and that, when querying an AST node, I should allways > use its API (is that usually #startPosition and #stopPosition?). > Me too The implication of my proposal is to also rename all the accessors so that the get proper names... which of course they do not have. Stef >> This is really annoying to have to look at the users of an instance >> variable to understand its role. > > Or have a well commented class to understand its role? It is not > allways possible to have instantaneously understandable instance > variable names. > > Thierry > > >

April 1, 2015

Re: [Pharo-dev] PharoV30.sources in Pharo 4.0 download

by Chris Cunningham

Is there a plan to condense changes and build a new sources file for the 4.0 release? Just curious -cbc On Wed, Apr 1, 2015 at 1:18 PM, Max Leske <maxleske(a)gmail.com> wrote: > > On 01 Apr 2015, at 22:12, Dmitri Zagidulin <dmitri(a)zagidulin.net> wrote: > > I noticed that the Pharo 4.0 full distribution zip file off of > Pharo.org/download contains PharoV30.sources (and not, > say, PharoV40.sources). > This is true for all 3 platforms: > - http://files.pharo.org/platform/Pharo4.0-mac.zip > - http://files.pharo.org/platform/Pharo4.0-win.zip > - http://files.pharo.org/platform/Pharo4.0-linux.zip > > Is there a typo in a CI job somewhere? > > > No, those are the latest sources :) The version simply tells you with what > image the sources have been created (I presume). Anyway, the version number > is not important. >

April 1, 2015

Re: [Pharo-dev] PharoV30.sources in Pharo 4.0 download

by Max Leske

> On 01 Apr 2015, at 22:12, Dmitri Zagidulin <dmitri(a)zagidulin.net> wrote: > > I noticed that the Pharo 4.0 full distribution zip file off of Pharo.org/download contains PharoV30.sources (and not, say, PharoV40.sources). > This is true for all 3 platforms: > - http://files.pharo.org/platform/Pharo4.0-mac.zip <http://files.pharo.org/platform/Pharo4.0-mac.zip> > - http://files.pharo.org/platform/Pharo4.0-win.zip <http://files.pharo.org/platform/Pharo4.0-win.zip> > - http://files.pharo.org/platform/Pharo4.0-linux.zip <http://files.pharo.org/platform/Pharo4.0-linux.zip> > > Is there a typo in a CI job somewhere? No, those are the latest sources :) The version simply tells you with what image the sources have been created (I presume). Anyway, the version number is not important.

April 1, 2015