December 2010 - Pharo-dev - lists.pharo.org

Re: [Pharo-project] WSDL, SOAP(HTTP) and XSD Schema (was Transforming WSDL in a local model)

by StÃ©phane Ducasse

Cedric this is excellent to see somebody pushing that. > > As a reminder, what I wanted was not to have something generic for all possible wsdl but to generate a kind of class model from the wsdl file of our engineering school planing application. This is kind of working now. But be aware this is a very very personal implementation not usable for others right now (but it might be in the future). OK learning and abstracting is the way to go :) Better something working than just talking :) > The wsdl (client) part I have for now might be useful for others but I doubt this is the case right now (needs to be more generic). Actually I started in the debugger with my particular wsdl files and then developed iteratively (so wsdl stuff can be fun too :) ) . > > During this process, I finally get more knowledge on wsdl "technology", so it might be doable to make it more generic. I have identified some model weaknesses in it that I need to fix (refactor) and then I'll try to reproduce the process with other wsdl files when I'll have some free time. There are plenty over there (http://www.xmethods.net/ve2/index.po) excellent. > For now, I'm more concerned on how to use SOAP and I'm just a bit stuck at the moment with the use of SOAP Opera (which I terribly hacked to make it kind of work). > > (1) I'd like to replace the use of KOM with ZnHTTPComponent > - authentication doesn't seem possible natively (but I might be wrong there) > -Soap Opera and Kom looks quite intermingled at the moment > > >> Sven: have you looked at SoapOpera ? Do you think it can be decoupled easyly from Kom to use ZnHttpComponent instead ? It would be good. This is good to have more clients of Zinc. > (2) the typing system is not implemented (I have a naive object wrapper on them - XSDSimpleType and XSDComplexType + XSDRestriction). > - I gave a look at Brenda Larcom implementation of XSD (X Files project on squeaksource - it's MIT even if not specified yet) > - I think I need a (clever) way to link XSDSimpleType/XSDComplexType with both my "wsdl" model and the soap implementation and of course map them back and forth with Smalltalk classes. > >> Any comments/idea here would be greatly appreciated ;-) - The "typing system" seems really important to me in term of interoperability with the outside world, but I'd like a clever way to have them mapped to smalltalk class/objects What do you need exactly do you have an example? > > Thanks in advance for your comments (to me I think we need, a SOAP decoupled from KOM and XSD stufs) > > Cheers, > > CÃ©drick > > > ps: the project is in squeaksource (iWSDL), but keep in mind this is a very very early stage... > >

Dec. 15, 2010

Re: [Pharo-project] Issue 3379 in pharo: ToolSet remove dependence to UIManager

by pharo＠googlecode.com

Updates: Labels: Milestone-1.3 Comment #1 on issue 3379 by stephane.ducasse: ToolSet remove dependence to UIManager http://code.google.com/p/pharo/issues/detail?id=3379 (No comment was entered for this change.)

Dec. 15, 2010

Re: [Pharo-project] Issue 3438 in pharo: SimpleMorphic and TextEditor

by pharo＠googlecode.com

Updates: Status: Closed Comment #2 on issue 3438 by stephane.ducasse: SimpleMorphic and TextEditor http://code.google.com/p/pharo/issues/detail?id=3438 in 12279

Dec. 15, 2010

Re: [Pharo-project] Issue 3437 in pharo: changeParagraph is sent in simple morphic and in tests

by pharo＠googlecode.com

Updates: Status: Closed Comment #2 on issue 3437 by stephane.ducasse: changeParagraph is sent in simple morphic and in tests http://code.google.com/p/pharo/issues/detail?id=3437 in 12279

Dec. 15, 2010

[Pharo-project] WSDL, SOAP(HTTP) and XSD Schema (was Transforming WSDL in a local model)

by CÃ©drick BÃ©ler

Hello all, Here are some news on my experiments on web services (driven by wsdl files) As a reminder, what I wanted was not to have something generic for all possible wsdl but to generate a kind of class model from the wsdl file of our engineering school planing application. This is kind of working now. But be aware this is a very very personal implementation not usable for others right now (but it might be in the future). The wsdl (client) part I have for now might be useful for others but I doubt this is the case right now (needs to be more generic). Actually I started in the debugger with my particular wsdl files and then developed iteratively (so wsdl stuff can be fun too :) ) . During this process, I finally get more knowledge on wsdl "technology", so it might be doable to make it more generic. I have identified some model weaknesses in it that I need to fix (refactor) and then I'll try to reproduce the process with other wsdl files when I'll have some free time. There are plenty over there (http://www.xmethods.net/ve2/index.po) For now, I'm more concerned on how to use SOAP and I'm just a bit stuck at the moment with the use of SOAP Opera (which I terribly hacked to make it kind of work). (1) I'd like to replace the use of KOM with ZnHTTPComponent - authentication doesn't seem possible natively (but I might be wrong there) -Soap Opera and Kom looks quite intermingled at the moment >> Sven: have you looked at SoapOpera ? Do you think it can be decoupled easyly from Kom to use ZnHttpComponent instead ? (2) the typing system is not implemented (I have a naive object wrapper on them - XSDSimpleType and XSDComplexType + XSDRestriction). - I gave a look at Brenda Larcom implementation of XSD (X Files project on squeaksource - it's MIT even if not specified yet) - I think I need a (clever) way to link XSDSimpleType/XSDComplexType with both my "wsdl" model and the soap implementation and of course map them back and forth with Smalltalk classes >> Any comments/idea here would be greatly appreciated ;-) - The "typing system" seems really important to me in term of interoperability with the outside world, but I'd like a clever way to have them mapped to smalltalk class/objects Thanks in advance for your comments (to me I think we need, a SOAP decoupled from KOM and XSD stufs) Cheers, CÃ©drick ps: the project is in squeaksource (iWSDL), but keep in mind this is a very very early stage...

Dec. 15, 2010

Re: [Pharo-project] New SqueakSource (was good news of the year)

by Luc Fabresse

<joke> graveyard.world.st just because once on disk, objects are just dead ;-) </joke> #Luc 2010/12/15 Nicolas Petton <petton.nicolas(a)gmail.com> > Le mardi 14 dÃ©cembre 2010 Ã 13:42 +0100, StÃ©phane Ducasse a Ã©crit : > > Smalltalksource is nice. > > I like it too. > > Nicolas > > >

Dec. 15, 2010

Re: [Pharo-project] [Vm-dev] A Git-based VM sources repositories for you

by Geoffroy Couprie

Hello, On Wed, Dec 15, 2010 at 8:25 AM, Igor Stasenko <siguctua(a)gmail.com> wrote: > > On 15 December 2010 03:35, David T. Lewis <lewis(a)mail.msen.com> wrote: >> >> On Tue, Dec 14, 2010 at 10:23:25PM +0100, Igor Stasenko wrote: >>> >>> Hello, community >>> >>> (preface) there was a discussion about using git for VM sources and >>> recently in pharo team, we are also discussed that, and decided to go >>> on. >>> So i started to look for an options and found that Gitorious is best >>> solution for us all. >>> In contrast to github, it doesn't charging users when their repository >>> grow beyond 300Mb in size. >>> And its totally free for open-source projects. >> >> Igor, >> >> Thanks for your initiative, and I expect a lively discussion moving forward ;) >> >> To the community at large, >> >> We have a wonderful mix of people participating in Squeak/Etoys/Phoro/Cuis >> and other projects, some of whom have be participating for many years, as well >> as many newer contributors. Regardless of how the VM(s) are managed in the >> future, I would ask that everyone with an interest in this please be respectful >> of the people who created the original work for our benefit. In particular, >> Andreas, Ian, John, and Eliot have contributed major VM development that >> they have made available to us with open source licenses, and Dan Ingalls' >> original portable VM is a remarkable piece of work from which we all are >> benefiting every day. >> >> So the meta-comment is: Let's all be respectful of the wishes of the original >> authors of these works. >> > > I am always acting with best intents for all of us. If creating a > separate code repository > goes against somebody's wishes, please speak. > I didn't want to cause any harm. > My intent is to make VM development process more open and flexible. > Thank you for taking care of this. Let's hope it will motivate more developers to study the code, learn about the VM and contribute tot his wonderful project!

Dec. 15, 2010

Re: [Pharo-project] Url comparison

by CÃ©drick BÃ©ler

> Isn't #printString intended for use in debugging? I used asString actually. cheers > > Peter > > On Fri, 10 Dec 2010 19:00:47 +0100, CÃ©drick BÃ©ler <cdrick65(a)gmail.com> wrote: > >> >> Le 10 dÃ©c. 2010 Ã 18:46, StÃ©phane Ducasse a Ã©crit : >> >>>> Do you think #= is needed for Url ? (ie. do I open an issue ?) >>> >>> yes. >> >> ok, I'll do that >> >>> Now I do not understand why you need to convert to text. >>> printString should be enough. >> >> true... :) >> >> >> Do I add >> >> Url >> hash >> >> self printstring hash >> >> >> >> >>> >>> >>>> >>>> if yes, is the following proposition enough ? >>>> or should it be more precise according to each subclass ? >>>> >>>> See you >>>> >>>> >>>>> >>>>> 'http://www.pharo-project.org' asUrl = 'http://www.pharo-project.org' asUrl. => false >>>>> >>>>> What about having: >>>>> >>>>> Url>> = anotherUrl >>>>> >>>>> ^self asText = anotherUrl asText >>>>> >>>>> Do we need to redefine hash too ? >>>>> >>>>> TIA >>>>> >>>>> CÃ©drick >>>>> >>>> >>>> >>> >>> >> >> >> > >

Dec. 15, 2010

Re: [Pharo-project] New SqueakSource (was good news of the year)

by Nicolas Petton

Dec. 15, 2010

Re: [Pharo-project] Url comparison

by Peter van Rooijen

Isn't #printString intended for use in debugging? Peter On Fri, 10 Dec 2010 19:00:47 +0100, CÃ©drick BÃ©ler <cdrick65(a)gmail.com> wrote: > > Le 10 dÃ©c. 2010 Ã 18:46, StÃ©phane Ducasse a Ã©crit : > >>> Do you think #= is needed for Url ? (ie. do I open an issue ?) >> >> yes. > > ok, I'll do that > >> Now I do not understand why you need to convert to text. >> printString should be enough. > > true... :) > > > Do I add > > Url >> hash > > self printstring hash > > > > >> >> >>> >>> if yes, is the following proposition enough ? >>> or should it be more precise according to each subclass ? >>> >>> See you >>> >>> >>>> >>>> 'http://www.pharo-project.org' asUrl = 'http://www.pharo-project.org' >>>> asUrl. => false >>>> >>>> What about having: >>>> >>>> Url>> = anotherUrl >>>> >>>> ^self asText = anotherUrl asText >>>> >>>> Do we need to redefine hash too ? >>>> >>>> TIA >>>> >>>> CÃ©drick >>>> >>> >>> >> >> > > >

Dec. 15, 2010