- Pharo-dev - lists.pharo.org

Re: [Pharo-project] need your attention: Package

by StÃ©phane Ducasse

Yes I would love that. I'm concerned about the migration from current situation to the new ones. Stef On Mar 19, 2011, at 11:28 AM, Marcus Denker wrote: > > On Mar 19, 2011, at 8:12 AM, StÃ©phane Ducasse wrote: > >>> >>> Is this a conscious decision to have an unique package per each >>> category name, or just a technical limitation? >> >> RPAckage has nothing to do with category matching: a package is a list of classes and methods. >> >> Now the problem is simply the following: >> >> you have a MC package FOO >> it contains FOO-Cat1 >> >> you load it: ok the loader could create >> RPAckage Foo >> and put FOO classes and Foo-Cat Classes in it >> >> Now you create a new category >> FOO-z what should I do >> add it to FOO >> create a package Foo-z >> >> We would like to get rid of the naming convention and matching on categories now we could have tags >> but tags should orthogonal to packages. >> > > > Imagine we would at the same time abandon both PackageInfo *and* System Categories. > > Then you would just do: > > > you have a MC package FOO > it contains FOO-Cat1 > > --> One RPackage Foo > > you have a MC package FOO > it contains FOO-Cat1 > FOO-Dog2 > > --> One RPackage Foo. We loose the sub structure. > > For the sub-structure, the way categories are now used in PackageInfo packages is > that they do not denote sub-packages, but just some convinient sorting without semantic > (to the language) meaning, just as method categories. > > So the perfect thing would be to have some form of ordering inside of RPackage (e.g. tagging). > > The case "Now you create a new Category" would not exist, as there are no categories anymore. > Categories should die. > > Marcus > > -- > Marcus Denker -- http://www.marcusdenker.de > INRIA Lille -- Nord Europe. Team RMoD. > >

March 19, 2011

Re: [Pharo-project] need your attention: Package

by Marcus Denker

On Mar 18, 2011, at 9:56 PM, StÃ©phane Ducasse wrote: > > - marcus was suggesting me to create a package with the same contents as the one of loaded by MC. > and to have tags to only represent categories. > > Now my time is short so I will > - probably not implement tags Even then, I would vote for mapping one PackageInfo --> one Rpackage. Categories are not packages, if we make every category a package we will end up in a big mess (as groups of these would have lots and lots of dependencies). Already the PackageInfo Packages for Network and Collection (the split into just the class categories - as- packages) was a mistake, purely driven by speed of MC. > - check again the implementation of RPackage and in particular the necessary compatibility layer, because I saw some strange > code. > - check the MC dependency on method category conventions, because some logic is not defined in the right place > like overrides in the MC tools and not in the PackageInfo > - check how a package gets created when loaded: the key question is that there is a problem to rely on categories to > associate classes to packages because we can end up with overlapping (normally the IDE captures the category renames > and change the packages accordingly). > - we should not rely on most-specific-category kind of pattern matching. > > So if you have suggestion please talk now. > > > Stef -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.

March 19, 2011

Re: [Pharo-project] need your attention: Package

by Marcus Denker

On Mar 19, 2011, at 8:12 AM, StÃ©phane Ducasse wrote: >> >> Is this a conscious decision to have an unique package per each >> category name, or just a technical limitation? > > RPAckage has nothing to do with category matching: a package is a list of classes and methods. > > Now the problem is simply the following: > > you have a MC package FOO > it contains FOO-Cat1 > > you load it: ok the loader could create > RPAckage Foo > and put FOO classes and Foo-Cat Classes in it > > Now you create a new category > FOO-z what should I do > add it to FOO > create a package Foo-z > > We would like to get rid of the naming convention and matching on categories now we could have tags > but tags should orthogonal to packages. > Imagine we would at the same time abandon both PackageInfo *and* System Categories. Then you would just do: you have a MC package FOO it contains FOO-Cat1 --> One RPackage Foo you have a MC package FOO it contains FOO-Cat1 FOO-Dog2 --> One RPackage Foo. We loose the sub structure. For the sub-structure, the way categories are now used in PackageInfo packages is that they do not denote sub-packages, but just some convinient sorting without semantic (to the language) meaning, just as method categories. So the perfect thing would be to have some form of ordering inside of RPackage (e.g. tagging). The case "Now you create a new Category" would not exist, as there are no categories anymore. Categories should die. Marcus -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.

March 19, 2011

Re: [Pharo-project] Method expressed as symbol, used as argument

by Mariano Martinez Peck

On Sat, Mar 19, 2011 at 2:40 AM, DougEdmunds <dougedmunds(a)gmail.com> wrote: > It was pointed out that given an instance method > MyClass >> doSomething > that > MyClass do: [:each | each doSomething]. > can be expressed as > MyClass do: #doSomething. > The reasons is this: check the #do: implementation and you will see all it does is to send #value:. But who says that only BlockClosure cna understand #value: ? Indeed, Symbol DOES UNDERSTAND value: ;) and it is: Symbol >> value: anObject ^anObject perform: self. So.....a selector is a Symbol.... Check this thread: http://forum.world.st/OffTopoc-Symbol-understands-value-nice-hehehehehhe-td… Now...whether it is cool to use it or not, it is another question. FOr iterating collections I would always use the normal approach. cheers Mariano > Can someone suggest something to read that explains the usage of # > here (#doSomething), the relation of methods to symbols, etc. > > Trying to wrap my head around it. Thanks. > > -dae > > > > -- > View this message in context: > http://forum.world.st/Method-expressed-as-symbol-used-as-argument-tp3388864… > Sent from the Pharo Smalltalk mailing list archive at Nabble.com. > >

March 19, 2011

Re: [Pharo-project] [call for help] Bugtracker

by Marcus Denker

On Mar 18, 2011, at 11:21 PM, Mariano Martinez Peck wrote: > > > Marcus, once I wrote this: http://www.pharo-project.org/community/issue-tracking > > but it seems people go directly to google code :( > It's not a problem to go directly and add a report... but it would be nice if people would care more about the bug issues they reported. If it's important enough to report, then it should be important enough to check the status once in a while... or just decided that indeed, it was not important. Marcus -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.

March 19, 2011

[Pharo-project] Equality and hash made easy

by Noury Bouraqadi

Hi, I proposed in early february an enhancement to avoid having redefine hash and = together and I would like to see it integrated in 1.3. My idea is that the implicit link between these two methods should be supported by the system. A first solution is proposed on the issue tracker http://code.google.com/p/pharo/issues/detail?id=3657 To make instances of some class comparable, simply make the class use the TEquality trait. TEquality relies on message valuesToCompareForEquality sent to objects for both = and hash. The default implmentation of this message is provided by Object. Override ONLY valuesToCompareForEquality method for custom equality check. 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

March 19, 2011

Re: [Pharo-project] Refactoring Browser Enhancements

by StÃ©phane Ducasse

Hi benjamin I published a new package (to revert the dependencies) in DirtyExperiments and a fixed the points raised by lukas. I added also some comments. I should fix testEnvironmentWrapper | printString wrapper | printString := BrowserEnvironment new referencesTo: #printString. wrapper := BrowserEnvironmentWrapper onEnvironment: printString. self assert: wrapper numberSelectors = printString numberSelectors. self assert: wrapper numberClasses = printString numberClasses. self assert: wrapper environment == printString Stef On Mar 18, 2011, at 6:39 PM, Benjamin wrote: > Hi lukas > > with Stef we refactored the Browser environment: > - now it can be parametrized by a system dictionary > - we also move selectionIn* to the AST-Core so that Environment can be loaded independently. > > Could you integrate the packages in rb? > There are > AST-Core > Refactoring-Critics > Refactoring-Environment > Refactoring-Test-Envrionment > > In DirtyExperiments on SqueakSource > > > Ben > > PS: All the test are green ;) > > > > >

March 19, 2011

Re: [Pharo-project] need your attention: Package

by StÃ©phane Ducasse

So I will check to see what can be done but it looks painful. Stef >> Is this a conscious decision to have an unique package per each >> category name, or just a technical limitation? > > RPAckage has nothing to do with category matching: a package is a list of classes and methods. > > Now the problem is simply the following: > > you have a MC package FOO > it contains FOO-Cat1 > > you load it: ok the loader could create > RPAckage Foo > and put FOO classes and Foo-Cat Classes in it > > Now you create a new category > FOO-z what should I do > add it to FOO > create a package Foo-z > > We would like to get rid of the naming convention and matching on categories now we could have tags > but tags should orthogonal to packages. > > >> I'd prefer to have a package which can allow an arbitrary category >> names in future. It may be that tools like browser are not prepared >> for that.. >> but not an internal information of package. To my thinking , classes >> which belong to package could have any category names.. >> Names should not mean anything.. it is just for humans. >> >>> >>> Now my time is short so I will >>> - probably not implement tags >>> - check again the implementation of RPackage and in particular the necessary compatibility layer, because I saw some strange >>> code. >>> - check the MC dependency on method category conventions, because some logic is not defined in the right place >>> like overrides in the MC tools and not in the PackageInfo >>> - check how a package gets created when loaded: the key question is that there is a problem to rely on categories to >>> associate classes to packages because we can end up with overlapping (normally the IDE captures the category renames >>> and change the packages accordingly). >>> - we should not rely on most-specific-category kind of pattern matching. >>> >> >> As i suggested, we should not rely on category naming in new packages at all. >> I think we could create an importer which using a legacy logic of >> making classes belong to some package based on their category name. > > Yes but this is a pain. :) > > >> In future that should be removed. >> >>> So if you have suggestion please talk now. >>> >>> >>> Stef >>> >> >> >> >> -- >> Best regards, >> Igor Stasenko AKA sig. >> > >

March 19, 2011

Re: [Pharo-project] Run a Pharo world on WebBrowser.

by Marcus Denker

On Mar 18, 2011, at 11:38 PM, nullPointer wrote: > I thinking do an application for run in desktop, browser and mobile devices. > I want know if is REALLY possible run Pharo in a web browser, like a > flash/java applet/silverlight application, without problems. Somebody has > proved that? How is difficult for the user install the mechanism in browser > for run the Pharo image? In the past, the way etoys was used was that you install the VM with a standard image as a plugin for the browser. Then .pr (Projects) where loaded just like much more interactive web pages. There where multiple problems with that a) It's a lot of work to keep the vm plugin working for real. A lot of time was invested from the side of the etoy team on that. b) Even though it's just "please install the plugin", people just don't do it. (partly because they are not allowed to, e.g. in a school setting) c) Projects... are problematic. Far too low level. Changes on the image lead to big problems if you want to be able to load old projects (as they are just a dumb of parts of the image... this will be especially fun as soon as you start to change real things in the system.) No idea what the latest state is.. but projects like Scratch decided it's easier to do a java (or flash?) based player app for the Scratch projects instead. There are two interesting developments: -> JavaScript. Millions are invested in JS VM technology. So one can start to do real smalltaks in JS. (Personally, I think having your own real VM, not one implemented ontop of another VM is preferable, though. Especially if you plan to do real strange things in the futuer as opposed to just wanting a smalltalk system in the browser.) -> Google Native Client. That one is very impressive. But it's unclear what googles plans are. We now they have no fear of promoting something and yet then kill it after people start to use it for real... Marcus -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.

March 19, 2011

Re: [Pharo-project] need your attention: Package

by StÃ©phane Ducasse

> > Is this a conscious decision to have an unique package per each > category name, or just a technical limitation? RPAckage has nothing to do with category matching: a package is a list of classes and methods. Now the problem is simply the following: you have a MC package FOO it contains FOO-Cat1 you load it: ok the loader could create RPAckage Foo and put FOO classes and Foo-Cat Classes in it Now you create a new category FOO-z what should I do add it to FOO create a package Foo-z We would like to get rid of the naming convention and matching on categories now we could have tags but tags should orthogonal to packages. > I'd prefer to have a package which can allow an arbitrary category > names in future. It may be that tools like browser are not prepared > for that.. > but not an internal information of package. To my thinking , classes > which belong to package could have any category names.. > Names should not mean anything.. it is just for humans. > >> >> Now my time is short so I will >> - probably not implement tags >> - check again the implementation of RPackage and in particular the necessary compatibility layer, because I saw some strange >> code. >> - check the MC dependency on method category conventions, because some logic is not defined in the right place >> like overrides in the MC tools and not in the PackageInfo >> - check how a package gets created when loaded: the key question is that there is a problem to rely on categories to >> associate classes to packages because we can end up with overlapping (normally the IDE captures the category renames >> and change the packages accordingly). >> - we should not rely on most-specific-category kind of pattern matching. >> > > As i suggested, we should not rely on category naming in new packages at all. > I think we could create an importer which using a legacy logic of > making classes belong to some package based on their category name. Yes but this is a pain. :) > In future that should be removed. > >> So if you have suggestion please talk now. >> >> >> Stef >> > > > > -- > Best regards, > Igor Stasenko AKA sig. >

March 19, 2011