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

Re: [Pharo-project] [squeak-dev] Re: usability of Pharo and Squeak

by GermÃ¡n Arduino

Hi Dale and folks: First of all let me say Dale, your capacity to deliver new projects is awesome! 2011/6/3 Dale Henrichs <dhenrich(a)vmware.com>: > > > ----- Original Message ----- > | From: "Andreas Wacknitz" <a.wacknitz(a)gmx.de> > | > | I hope you will be successful. I don't like programming web > | interfaces because I am still not convinced that they are superior > | to traditional GUI's. > | Especially I don't like the mix of HMTL, JavaScript and CSS, even if > | you can replace HTML by Smalltalk with Seaside. > | And I also need to be convinced that Seaside is really such a big > | thing. I haven't seen advanced GUI's in Seaside yet. > | The examples I have seen in the Seaside book and the Seaside site are > | all simplistic. > | Maybe tODE will change my mind :D > > Thanks Andreas, > > For advanced Seaside apps, check out NeXTPLAN[1]... > > As for tODE, the framework/approach isn't directly tied to Seaside ... I think tODE could be done in Morphic, but a Morphic-based tODE wouldn't serve my needs. Not only does GemStone lack a native GUI, but the GLASS applications are running on machines in a server farm somewhere and native GUIs aren't very responsive over the WAN, whereas serving pages from a remote server can provide acceptable performance. > This is a great explanation. Lot of times I must explain myself (as a freelancer working for customers around the world) why I select web interfaces, this is the reason. Working with different datacenters, mix of networks and servers need to have a fast way to interact with them, and is the reason because the web is here. > I do agree that I seem to be working with an unholy mixture of HTML, Javascript, and CSS, but Seaside does help hide the ugliness of HTML and Javascript... > I'm not a javascript guy neither, but like or not, javascript is the engine that energize the web in our days, this was the main reason becuase Diego GD developed the ST2JS [1] some years ago, to program Javscript using Smalltalk. Unfortunately I had not more time, nor a way to insert in a payed project to continue this development, but I'm convinced that is a very very goof helper tool for Smalltalkers doing web development. GermÃ¡n. [1] ST2JS stand for "Smalltalk to Javascript", is part of SWT (Smalltalk Web Toolkint), a tiny reference here: http://www.slideshare.net/garduino1/swt-2009

June 4, 2011

Re: [Pharo-project] Compiling and using the FT2Plugin

by Igor Stasenko

On 4 June 2011 11:34, Alain Plantec <alain.plantec(a)yahoo.com> wrote: > Hi Igor, > > #configureFT2Plugin: is the place for #addExternalLibraries: but it only > works for FT2Plugin compiled as internal plugin. > To compile the FT2Plugin as external, you also have to use > #addExternalLibraries: #('freetype') but not sent from #configureFT2Plugin:. > It is because of cmake and where #configurePlugin:with: is sent from the > #generate method. > in the #generate method, #configurePlugin:with: is sent Â **BEFORE** the > 'add_library' cmd production for cmake. > it is ok for internal plugin. but, if the plugin is external, then, > #addExternalLibraries: #('freetype') > has to be invoked **AFTER** the 'add_library' cmd production for cmake. > > I've implemented a #postConfigurePlugin:with:. same logic as > #configurePlugin:with: but sent from > the generate method after the 'add_library' cmd production. > > generate > ... > Â Â vmGen config configurePlugin: plugin with: self. > ... > Â Â internal > Â Â Â Â ifTrue: [self cmd: 'add_library' params: Â name , ' STATIC > ${sources}'.] > Â Â Â Â ifFalse: [self cmd: 'add_library' params: name , ' SHARED > ${sources}']. > > Â Â vmGen config postConfigurePlugin: plugin with: self. Â "added post > configuration here" > ... > Yeah. I am also stumbled with this couple of times. The workaround would be to use variables, so in #addExternalLibrary , it should set (or better append ) library to the list, and then generator will simply use this list in a single command for linking with target. > I will open an issue and put all my stuffs there. > good. thanks. > Cheers > Alain > > -- Best regards, Igor Stasenko AKA sig.

June 4, 2011

[Pharo-project] Call For Papers: IWST@ESUG2011

by Alain Plantec

IWST @ ESUG 2011 International Workshop on Smalltalk Technologies ACM In-Cooperation August 23th, 2011 Edinburgh, Scotland http://www.esug.org/Conferences/2011/International-Workshop-on-Smalltalk-Te… SUBMISSION DEADLINE : June 17, 2011 Authors of the best accepted papers will be invited to submit extended versions of their papers for publication in a special issue of 'Software: Practice and Experience' ------------------------------------------------------------------------ ESUG 2011 Smalltalk joint event IMPORTANT DATES * Submission deadline: June 17, 2011 * Notification deadline: July 15, 2011 GOALS AND SCOPE We are proud to invite submisssions to the International Workshop on Smalltalk Technologies which is held as part of the ESUG 2011 joint event at Edinburgh. The goals of the workshop is to create a forum around advances or experience in Smalltalk and to trigger discussions and exchanges of ideas. Participants are invited to submit short and not-so short research articles. We will not enforce any length restriction. However we expect papers of two kinds: * Short position papers describing emerging ideas. * Long research papers with deeper description of experiments and of research results. TOPICS We welcome contributions on all aspects, theoretical as well as practical, of Smalltalk related topics such as: * Aspect-oriented programming, * Meta-programming and Meta-modeling, * Frameworks, * Interaction with other languages, * Implementation, new dialects or languages implemented in Smalltalk, * Tools, * Design patterns, * Experience reports PUBLICATION Both submissions and final papers must be prepared using the ACM SIGPLAN 10 point format. Templates for Word and LaTeX are available at http://www.acm.org/sigs/sigplan/authorInformation.htm; this site also contains links to useful informations on how to write effective submissions. PROGRAM CHAIRS: * LoÃ¯c Lagadec (Labsticc, University of Brest, France) * Alain Plantec (Lisyc, University of Brest, France) PROGRAM COMMITTEE: * Gabriela Arevalo * Marcus Denker * Johan Fabry * Lukas Renggli * Hernan Wilkinson * Damien Cassou * Alexandre Bergel * Jordi Delgado * Johan Brichau

June 4, 2011

Re: [Pharo-project] Problem with Traits and #initialize.

by Lukas Renggli

>> What you observe is what is implemented in these methods. Normally you >> should never call class-side initialize methods manually, the system >> does > > yeah, the system *should* do it for me. I cannot reproduce it, but I find > myself several times having to explicitly send class side #initialize > becasue Monticello did not do it. Monticello sends class side #initialize whenever it loads a new version of that method, that is on initial load and/or when it updates a method after load/merge. It does not send #initialize in any other case and it does not send super #initialize methods (that would break the system in most cases). Note that this has exactly the same semantics as with change-sets: If a change-set includes a class-side initialize method, it is executed at the end of the file-in. > ((PackageInfo named: 'XXX') classes select: [:each | each isBehavior ]) do: [:each | each initialize] Do Gofer new package: 'XXX'; reinitialize It does the "right" thing. Lukas -- Lukas Renggli www.lukas-renggli.ch

June 4, 2011

Re: [Pharo-project] [Pharo-users] we should fix the timeout because I have problem to commit morphic changes.

by Mariano Martinez Peck

I was bit for that today because SS was slow... And yes, I think it may be related to Zinc. On Sat, Jun 4, 2011 at 12:18 PM, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr > wrote: > hi sven/marcus > > what is the solution for the timeout? > because I get in trouble and cannot publish morphic or I have to retry and > this is not good. > > Stef > > > > -- Mariano http://marianopeck.wordpress.com

June 4, 2011

Re: [Pharo-project] Problem with Traits and #initialize.

by Mariano Martinez Peck

On Sat, Jun 4, 2011 at 12:28 PM, Lukas Renggli <renggli(a)gmail.com> wrote: > Did you look at TraitBehavior>>#initialize and Behavior>>#initialize? > They both clear the method dictionary, reset superclass, clear users > and traits, etc. > > Thanks for the point. I can see now. > What you observe is what is implemented in these methods. Normally you > should never call class-side initialize methods manually, the system > does yeah, the system *should* do it for me. I cannot reproduce it, but I find myself several times having to explicitly send class side #initialize becasue Monticello did not do it. > this for you if necessary. > > Lukas > > On 4 June 2011 12:04, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: > > why would it make sense to send initialize? > > > > Stef > > > > On Jun 4, 2011, at 11:42 AM, Mariano Martinez Peck wrote: > > > >> 1) Browse the example Trait1 and you will see methods like: > >> Trait1 >> c > >> > >> ^ 'Trait1>>c' > >> > >> and: > >> > >> Trait1 >> c1 > >> > >> ^ 'Trait1>>c1' > >> > >> > >> 2) Now, execute "Trait1 initialize" > >> 3) BOOM. Trait1 has no longo any method. In the browse they are shown > like: > >> > >> Trait1 >> c > >> "This method does not exist." > >> self halt. > >> > >> Trait2 >> c1 > >> "This method does not exist." > >> self halt. > >> > >> .... > >> > >> If I see Trait >> initialize > >> super initialize. > >> classTrait := ClassTrait for: self > >> > >> > >> and then ClassTrait for: > >> "This method does not exist." > >> self halt. > >> > >> > >> wtf ?? that method is removed? this even happens before doing Trait1 > initialize. > >> > >> so...any idea? > >> > >> > >> -- > >> Mariano > >> http://marianopeck.wordpress.com > >> > > > > > > > > > > -- > Lukas Renggli > www.lukas-renggli.ch > > -- Mariano http://marianopeck.wordpress.com

June 4, 2011

Re: [Pharo-project] Problem with Traits and #initialize.

by Mariano Martinez Peck

On Sat, Jun 4, 2011 at 12:04 PM, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr > wrote: > why would it make sense to send initialize? > Because the system has fucking names. PackageInfo >> classes answers both, traits and classes. I wanted to send #initialize to all classes of a package. Of course, I did a workaround: ((PackageInfo named: 'XXX') classes select: [:each | each isBehavior ]) do: [:each | each initialize] Funny, I reported exactly the same 1 year ago: http://forum.world.st/Bug-PackageInfo-classes-return-also-Traits-tp2165137p… what a bad memory I have > > Stef > > On Jun 4, 2011, at 11:42 AM, Mariano Martinez Peck wrote: > > > 1) Browse the example Trait1 and you will see methods like: > > Trait1 >> c > > > > ^ 'Trait1>>c' > > > > and: > > > > Trait1 >> c1 > > > > ^ 'Trait1>>c1' > > > > > > 2) Now, execute "Trait1 initialize" > > 3) BOOM. Trait1 has no longo any method. In the browse they are shown > like: > > > > Trait1 >> c > > "This method does not exist." > > self halt. > > > > Trait2 >> c1 > > "This method does not exist." > > self halt. > > > > .... > > > > If I see Trait >> initialize > > super initialize. > > classTrait := ClassTrait for: self > > > > > > and then ClassTrait for: > > "This method does not exist." > > self halt. > > > > > > wtf ?? that method is removed? this even happens before doing Trait1 > initialize. > > > > so...any idea? > > > > > > -- > > Mariano > > http://marianopeck.wordpress.com > > > > > -- Mariano http://marianopeck.wordpress.com

June 4, 2011

Re: [Pharo-project] Problem with Traits and #initialize.

by Lukas Renggli

Did you look at TraitBehavior>>#initialize and Behavior>>#initialize? They both clear the method dictionary, reset superclass, clear users and traits, etc. What you observe is what is implemented in these methods. Normally you should never call class-side initialize methods manually, the system does this for you if necessary. Lukas On 4 June 2011 12:04, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: > why would it make sense to send initialize? > > Stef > > On Jun 4, 2011, at 11:42 AM, Mariano Martinez Peck wrote: > >> 1) Browse the example Trait1 Â and you will see methods like: >> Trait1 >> c >> >> Â Â ^ 'Trait1>>c' >> >> and: >> >> Trait1 >> c1 >> >> Â Â ^ 'Trait1>>c1' >> >> >> 2) Now, execute "Trait1 initialize" >> 3) BOOM. Trait1 has no longo any method. In the browse they are shown like: >> >> Trait1 >> c >> Â Â "This method does not exist." >> Â Â self halt. >> >> Trait2 >> c1 >> Â Â "This method does not exist." >> Â Â self halt. >> >> .... >> >> If I see Trait >> initialize >> Â Â super initialize. >> Â Â classTrait := ClassTrait for: self >> >> >> and then ClassTrait for: >> Â Â "This method does not exist." >> Â Â self halt. >> >> >> wtf ?? Â that method is removed? this even happens before doing Trait1 initialize. >> >> so...any idea? >> >> >> -- >> Mariano >> http://marianopeck.wordpress.com >> > > > -- Lukas Renggli www.lukas-renggli.ch

June 4, 2011

Re: [Pharo-project] Problem with Traits and #initialize.

by StÃ©phane Ducasse

why would it make sense to send initialize? Stef On Jun 4, 2011, at 11:42 AM, Mariano Martinez Peck wrote: > 1) Browse the example Trait1 and you will see methods like: > Trait1 >> c > > ^ 'Trait1>>c' > > and: > > Trait1 >> c1 > > ^ 'Trait1>>c1' > > > 2) Now, execute "Trait1 initialize" > 3) BOOM. Trait1 has no longo any method. In the browse they are shown like: > > Trait1 >> c > "This method does not exist." > self halt. > > Trait2 >> c1 > "This method does not exist." > self halt. > > .... > > If I see Trait >> initialize > super initialize. > classTrait := ClassTrait for: self > > > and then ClassTrait for: > "This method does not exist." > self halt. > > > wtf ?? that method is removed? this even happens before doing Trait1 initialize. > > so...any idea? > > > -- > Mariano > http://marianopeck.wordpress.com >

June 4, 2011

Re: [Pharo-project] loading code from command line inovcation

by Marcus Denker

On Jun 3, 2011, at 10:57 PM, Hilaire Fernandes wrote: > What am I suppose to understand from this url? > a better link is http://code.google.com/p/pharo/issues/detail?id=4319 > Le 31/05/2011 10:27, StÃ©phane Ducasse a Ã©crit : >> http://code.google.com/p/pharo/issues/list?thanks=4319 >> >> >> On May 30, 2011, at 9:12 PM, Hilaire Fernandes wrote: >> >>> Some more tests: >>> >>> 1. this crash the image: >>> squeak PharoCore-1.3.image `pwd`/cacaboudin >>> >>> cacaboudin does not exist >>> >>> 2. this crash the image: >>> squeak PharoCore-1.3.image `pwd`/empty.st >>> >>> where empty.st is empty >>> >>> 3. this work: >>> >>> squeak PharoCore-1.3.image `pwd`/test.st >>> >>> where test.st only contains: Transcript show: 'hello' >>> >>> >>> >>> >>> >>> >>> >>> Le 28/05/2011 22:53, Hilaire Fernandes a Ã©crit : >>>> Still does not work >>>> >>>> Hilaire >>>> >>>> Le 28/05/2011 21:02, Guillermo Polito a Ã©crit : >>>>> It seems code loader is not sending #startRetrieval to the UrlDownload >>>>> Request, and the Semaphore waits forever. >>>>> >>>>> I've put a changeset with what I think is a fix in the tracker :). >>>>> >>>>> On Sat, May 28, 2011 at 2:27 PM, Hilaire Fernandes >>>>> <hilaire.fernandes(a)gmail.com >>>>> <mailto:hilaire.fernandes@gmail.com>> wrote: >>>>> >>>>> And the debug is there. >>>>> The halt does not crash the VM in normal condition. (ie Alt-[.] is ok) >>>>> >>>>> Hilaire >>>>> >>>>> >>>>> Le 28/05/2011 19:07, Hilaire Fernandes a Ã©crit : >>>>>> And when there are no halt in commandLineHandlerAction:, the image is >>>>>> frozen >>>>>> >>>>>> Le 28/05/2011 18:53, Hilaire Fernandes a Ã©crit : >>>>>>> With latest pahro core 1.3, I can't load script with command like >>>>>>> >>>>>>> >>>>>>> squeak pharo "`pwd`\myScript.st" >>>>>>> >>>>>>> When I put a halt at the begining of commandLineHandlerAction: >>>>>>> aCommandLine the VM crash >>>>>>> >>>>>>> VM is 4.0.3-2202 >>>>>>> >>>>>>> Hilaire >>>>>>> >>>>>>> >>>>>> >>>>>> >>>>> >>>>> >>>>> -- >>>>> Education 0.2 -- http://blog.ofset.org/hilaire >>>>> >>>>> >>>> >>>> >>> >>> >>> -- >>> Education 0.2 -- http://blog.ofset.org/hilaire >>> >>> >> >> >> > > > -- > Education 0.2 -- http://blog.ofset.org/hilaire > > -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.

June 4, 2011