August 2015 - Pharo-users - lists.pharo.org

Re: [Pharo-users] ***UNCHECKED*** Re: [ANN] Magma 1.5
by Pablo R. Digonzelli Aug. 18, 2015

Aug. 18, 2015

Yes I checkesd. It is by far better approach to pharo and I like it very much. The problem i found was when i stress mongo, it lost some transactions (i.e 10000 rewrites lost about 80 or more don't remember exactly) . In bussiness apps i cannot take that risk. Ing. Pablo Digonzelli Software Solutions IP-Solutiones SRL Metrotec SRL 25 de Mayo 521 San Miguel de TucumÃ¡n Email: pdigonzelli(a)softsargentina.com pdigonzelli(a)gmail.com Cel: 5493815982714 ----- Mensaje original ----- De: "Esteban Lorenzano" <estebanlm(a)gmail.com> Para: "Any question about pharo is welcome" <pharo-users(a)lists.pharo.org> CC: stephan(a)stack.nl, "Torsten Bergmann" <astares(a)gmx.de>, hilaire(a)drgeo.eu Enviados: Martes, 18 de Agosto 2015 12:45:05 Asunto: ***UNCHECKED*** Re: [Pharo-users] [ANN] Magma 1.5 > On 18 Aug 2015, at 17:20, Pablo R. Digonzelli <pdigonzelli(a)gmail.com> wrote: > > I can colaborate. I do not know if i have the time and technical skills to do it alone, but for shure i can colaborate. > I think magma will be a very interesting option for object persistence in pharo. > I am currently working with garage and glorp and i suffers the innecesary complexity and other bad things thet persistence layer introduce in developing pharo apps. have you checked voyage? (to work with a MongoDB)â¦ it will avoid all the pain of using a relational database. Esteban ps: But yes, I do believe having magma would be cool too. > > Pharo > > > Ing. Pablo Digonzelli > Software Solutions > IP-Solutiones SRL > Metrotec SRL > 25 de Mayo 521 > San Miguel de TucumÃ¡n > Email: pdigonzelli(a)softsargentina.com > pdigonzelli(a)gmail.com > Cel: 5493815982714 > > ----- Mensaje original ----- > De: "Torsten Bergmann" <astares(a)gmx.de> > Para: "Any question about pharo is welcome" <pharo-users(a)lists.pharo.org> > CC: stephan(a)stack.nl, stepharo(a)free.fr, hilaire(a)drgeo.eu, "Pablo Digonzelli" <pdigonzelli(a)gmail.com> > Enviados: Martes, 18 de Agosto 2015 10:26:44 > Asunto: Fwd: [ANN] Magma 1.5 > > Stephan Eggermont wrote: >> It is possible, and it needs someone who can deal with low-level >> details. Magma has its own mailing list. Chris has repeatedly said he'd >> be willing to help out, but is not willing to be the Pharo port maintainer. > > I asked Chris directly and you can find his answer here on the magma list: > > http://lists.squeakfoundation.org/pipermail/magma/2015-August/001955.html > > As ever it requires someone to move this forward to make it happen. > But as there was an initiative with Magma 1.3 for Pharo 1.3 already it does not > seem impossible. > > Also - as Stef already pointed out - community services like CI could be used > to maintain it once an initial up to date port was done. > > Bye > T. >

3 2

Re: [Pharo-users] [ANN] Magma 1.5
by Pablo R. Digonzelli Aug. 18, 2015

Aug. 18, 2015

I can colaborate. I do not know if i have the time and technical skills to do it alone, but for shure i can colaborate. I think magma will be a very interesting option for object persistence in pharo. I am currently working with garage and glorp and i suffers the innecesary complexity and other bad things thet persistence layer introduce in developing pharo apps. Pharo Ing. Pablo Digonzelli Software Solutions IP-Solutiones SRL Metrotec SRL 25 de Mayo 521 San Miguel de TucumÃ¡n Email: pdigonzelli(a)softsargentina.com pdigonzelli(a)gmail.com Cel: 5493815982714 ----- Mensaje original ----- De: "Torsten Bergmann" <astares(a)gmx.de> Para: "Any question about pharo is welcome" <pharo-users(a)lists.pharo.org> CC: stephan(a)stack.nl, stepharo(a)free.fr, hilaire(a)drgeo.eu, "Pablo Digonzelli" <pdigonzelli(a)gmail.com> Enviados: Martes, 18 de Agosto 2015 10:26:44 Asunto: Fwd: [ANN] Magma 1.5 Stephan Eggermont wrote: >It is possible, and it needs someone who can deal with low-level >details. Magma has its own mailing list. Chris has repeatedly said he'd >be willing to help out, but is not willing to be the Pharo port maintainer. I asked Chris directly and you can find his answer here on the magma list: http://lists.squeakfoundation.org/pipermail/magma/2015-August/001955.html As ever it requires someone to move this forward to make it happen. But as there was an initiative with Magma 1.3 for Pharo 1.3 already it does not seem impossible. Also - as Stef already pointed out - community services like CI could be used to maintain it once an initial up to date port was done. Bye T.

2 1

ClassCategoryReader and ClassCommentReader
by Guillermo Polito Aug. 18, 2015

Aug. 18, 2015

These two classes seem to be unused. They implement the #scanFor: method but the only senders of #scanFor: send it explicitly to RunArray, like: RunArray scanFor: ... Does anybody use it? Because it looks like an old class that could be removed. Guille

2 2

Fwd: [ANN] Magma 1.5
by Torsten Bergmann Aug. 18, 2015

Aug. 18, 2015

Stephan Eggermont wrote: >It is possible, and it needs someone who can deal with low-level >details. Magma has its own mailing list. Chris has repeatedly said he'd >be willing to help out, but is not willing to be the Pharo port maintainer. I asked Chris directly and you can find his answer here on the magma list: http://lists.squeakfoundation.org/pipermail/magma/2015-August/001955.html As ever it requires someone to move this forward to make it happen. But as there was an initiative with Magma 1.3 for Pharo 1.3 already it does not seem impossible. Also - as Stef already pointed out - community services like CI could be used to maintain it once an initial up to date port was done. Bye T.

1 0

Smalltalker required
by S Krish Aug. 18, 2015

Aug. 18, 2015

" We are looking for smalltalk developers in US. Must have L1 or H1, or any other valid work visa. Job location would be Atlanta. JD: 1.) Visual age Smalltalk V5.0 2.) Strong debugging, File handling and UI designing using VAST5.0 3.) Strong database skillset using SQL 4.) Analyzing production support tickets 5.) Communication protocols like Web services, OLE, DDE, TCP-IP Please send profiles to ravinder.phogat(a)zensar.in" <ravinder.phogat(a)zensar.in>

2 1

Fwd: [ANN] Magma 1.5
by Pablo R. Digonzelli Aug. 17, 2015

Aug. 17, 2015

FYI. Is it interesting for pharo comunity migrate magma from squeak to pharo?. Anyone knows if it is possible to do? Ing. Pablo Digonzelli Software Solutions IP-Solutiones SRL Metrotec SRL 25 de Mayo 521 San Miguel de TucumÃ¡n Email: pdigonzelli(a)softsargentina.com pdigonzelli(a)gmail.com Cel: 5493815982714 ----- Mensaje reenviado ----- De: "Chris Muller" <ma.chris.m(a)gmail.com> Para: "magma" <magma(a)lists.squeakfoundation.org> Enviados: Jueves, 13 de Agosto 2015 12:42:16 Asunto: Fwd: [ANN] Magma 1.5 I am pleased to announce version 1.5 of Magma for Squeak 5, now available on SqueakMap. Magma allows multiple Squeak images to collaborate on a single, large object model, with the robustness and control expected from a database. It offers the most transparent db access possible for Smalltalk, affording the user the ability to develop complex, performant designs, iteratively, on-the-fly. It has been designed for "continuous flow" development, the way Smalltalkers like and expect to work. For example, I could have connections open to 3 separate databases, open transactions in any of them, and having restructured a class hierarchy in the model, and stepping through the debugger when that "final boarding call" for my flight is announced. Thanks to the image, this scenario has never been a problem for Smalltalkers and Magma is deliberate to ensure this flow is maintained. Once at 10K feet, I can resume stepping through that same debugger within 5 seconds of restarting the image, DB connections intact, commit my transactions when I'm ready, done. Magma handles every aspect of that use-case correctly even in multi-user environments, and has so many safety and integrity features, it is the safest way to develop and keep a model in Squeak. This release coincides with the release of Squeak 5, and has many improvements and fixes over Magma 1.4. Detailed notes about these improvements are available at http://wiki.squeak.org/squeak/6209. - Chris _______________________________________________ Magma mailing list Magma(a)lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/magma

4 3

how to increase memory for the jvm through jniport?
by Tudor Girba Aug. 17, 2015

Aug. 17, 2015

Hi, When using JNIPort I am getting a "java.lang.OutOfMemory: Java heap space". To go around this, I would need to provide more memory to the JVM. Essentially, I would like to provide the correspondent of a command line like: java -Xmx4000m How should I do this? Cheers, Doru -- www.tudorgirba.com "Every thing has its own flow"

2 2

how to reinitialize the jvm in jniport?
by Tudor Girba Aug. 17, 2015

Aug. 17, 2015

Hi, I am trying to shutdown a Jvm instance but when I try to use a new instance, I get a "JNIError: unknown error". Here is what I do on OS X using Pharo 3.0: Gofer new smalltalkhubUser: 'JNIPort' project: 'JNIPort'; configuration; loadVersion: '3.0-baseline'. JNIPortJNIInterface libraryFile: '/System/Library/Java/JavaVirtualMachines/1.6.0.jdk/Contents/Libraries/libclient.dylib'. jvm := JVM newWithDefaultSettings. jvm findClass: 'java.lang.String'. "==> public final java.lang.String Instance class: JavaLangString Static class: StaticJavaLangObject" jvm shutdown. "with or without this line, the below error is the same" anotherJvm := JVM newWithDefaultSettings. ==> "JNIError: unknown error" What am I doing wrong? Cheers, Doru -- www.tudorgirba.com "Every thing has its own flow"

2 2

removeFromChanges removeFromSystemUnlogged
by Guillermo Polito Aug. 17, 2015

Aug. 17, 2015

There are many senders of these two messages, that look like the following. Most of them are tests. cleaningResources Smalltalk globals at: #ManifestManifestResourcesTests ifPresent: [ :cl | cl removeFromChanges; removeFromSystemUnlogged ] Shouldn't be cleaner to just create the classes silently? Guille

1 0

Mongo lost references - bug?
by Sabine KnÃ¶fel Aug. 17, 2015

Aug. 17, 2015

Hi Esteban, I found a problem in mongo/voyage and I have reduced and extracted it for repro. I ask for your opinion if this is a bug. If you file in the code below, the method >>testDemo creates a simple model: Paper has 1:1 reference to Planet. Paper has 1:1 reference to Star. Star has 1:N reference to Planet. Star has 1:1 reference to Paper. Planet has 1:1 reference to Paper. If you run >>testDemo, you see that the model which was created is ok with all its references. (self assert: thePaper planet = thePaper star planets first.) ==> ok If you reset the repository and load Paper from database with Paper select all and make the same check, the references are lost. VORepository current reset. theReloadedPaper := Paper selectAll first. self assert: theReloadedPaper planet = theReloadedPaper star planets first ==>not ok But if you remove the comment within >>mongoPlanets which leads to sending a message to the newly loaded object, then it works. write: [ :star :planets | star planets: (planets "select: [:each | each someMethod]") ]); Do I have to send a message to each of the instances after loading? I think no. If I did not understand correctly how to use voyage, please tell me. Regards Sabine Object subclass: #Planet instanceVariableNames: 'planetname star paper' classVariableNames: '' poolDictionaries: '' category: 'RKA24-Demo'! !Planet methodsFor: 'accessing' stamp: 'sabineknoefel 12/2/2013 12:05'! paper ^ paper! ! !Planet methodsFor: 'accessing' stamp: 'sabineknoefel 12/2/2013 12:05'! paper: anObject paper := anObject! ! !Planet methodsFor: 'accessing' stamp: 'sabineknoefel 7/18/2013 16:17'! planetname ^ planetname! ! !Planet methodsFor: 'accessing' stamp: 'sabineknoefel 7/18/2013 16:17'! planetname: anObject planetname := anObject! ! !Planet methodsFor: 'accessing' stamp: 'sabineknoefel 12/2/2013 11:53'! star ^ star! ! !Planet methodsFor: 'accessing' stamp: 'sabineknoefel 12/2/2013 11:53'! star: anObject star := anObject! ! !Planet methodsFor: 'testing' stamp: 'sabineknoefel 12/2/2013 12:25'! someMethod ^true! ! "-- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- "! Planet class instanceVariableNames: ''! !Planet class methodsFor: 'as yet unclassified' stamp: 'sabineknoefel 7/18/2013 16:16'! isVoyageRoot ^true! ! !Planet class methodsFor: 'as yet unclassified' stamp: 'sabineknoefel 12/2/2013 12:23'! mongoStar <mongoDescription> ^ VOMongoToOneDescription new attributeName: 'star'; kind: Star; accessor: (MAPluggableAccessor read: [ :planet | planet star ] write: [ :planet :star | planet star: star ]); yourself! ! !Planet class methodsFor: 'as yet unclassified' stamp: 'sabineknoefel 12/2/2013 12:30'! testDemo "self testDemo" | thePlanet theStar thePaper theReloadedPaper | "create simple model" thePaper := Paper new theme: 'a paper about 42'. thePlanet := Planet new planetname: 'Orion'; paper: thePaper. theStar := Star new starname: 'Betelgeuse'; paper: thePaper. theStar addPlanet: thePlanet. thePaper star: theStar. thePaper planet: thePlanet. thePaper save. "check references - ok" self assert: thePaper planet = thePaper star planets first. VORepository current reset. theReloadedPaper := Paper selectAll first. "references are lost now" self assert: theReloadedPaper planet = theReloadedPaper star planets first! ! Object subclass: #Star instanceVariableNames: 'starname planets paper' classVariableNames: '' poolDictionaries: '' category: 'RKA24-Demo'! !Star methodsFor: 'api' stamp: 'sabineknoefel 7/18/2013 16:21'! addPlanet: aPlanet self planets add: aPlanet. aPlanet star: self! ! !Star methodsFor: 'accessing' stamp: 'sabineknoefel 12/2/2013 12:05'! paper ^ paper! ! !Star methodsFor: 'accessing' stamp: 'sabineknoefel 12/2/2013 12:05'! paper: anObject paper := anObject! ! !Star methodsFor: 'accessing' stamp: 'sabineknoefel 7/18/2013 16:20'! planets planets ifNil: [ self planets: OrderedCollection new ]. ^ planets! ! !Star methodsFor: 'accessing' stamp: 'sabineknoefel 7/18/2013 16:20'! planets: anObject planets := anObject! ! !Star methodsFor: 'accessing' stamp: 'sabineknoefel 7/18/2013 16:18'! starname ^ starname! ! !Star methodsFor: 'accessing' stamp: 'sabineknoefel 7/18/2013 16:18'! starname: anObject starname := anObject! ! "-- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- "! Star class instanceVariableNames: ''! !Star class methodsFor: 'as yet unclassified' stamp: 'sabineknoefel 7/18/2013 16:16'! isVoyageRoot ^true! ! !Star class methodsFor: 'as yet unclassified' stamp: 'sabineknoefel 12/2/2013 12:31'! mongoPlanets "remove comment below then references are not lost anymore" <mongoDescription> ^ VOMongoToManyDescription new attributeName: 'planets'; kind: Planet; accessor: (MAPluggableAccessor read: [ :star | star planets ] write: [ :star :planets | star planets: (planets "select: [:each | each someMethod]") ]); yourself ! ! Object subclass: #Paper instanceVariableNames: 'star planet theme' classVariableNames: '' poolDictionaries: '' category: 'RKA24-Demo'! !Paper methodsFor: 'accessing' stamp: 'sabineknoefel 12/2/2013 11:54'! planet ^ planet! ! !Paper methodsFor: 'accessing' stamp: 'sabineknoefel 12/2/2013 11:54'! planet: anObject planet := anObject! ! !Paper methodsFor: 'accessing' stamp: 'sabineknoefel 12/2/2013 11:54'! star ^ star! ! !Paper methodsFor: 'accessing' stamp: 'sabineknoefel 12/2/2013 11:54'! star: anObject star := anObject! ! !Paper methodsFor: 'accessing' stamp: 'sabineknoefel 12/2/2013 11:54'! theme ^ theme! ! !Paper methodsFor: 'accessing' stamp: 'sabineknoefel 12/2/2013 11:54'! theme: anObject theme := anObject! ! "-- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- "! Paper class instanceVariableNames: ''! !Paper class methodsFor: 'as yet unclassified' stamp: 'sabineknoefel 12/2/2013 11:48'! isVoyageRoot ^true! ! !Paper class methodsFor: 'as yet unclassified' stamp: 'sabineknoefel 12/2/2013 12:23'! mongoPlanet <mongoDescription> ^ VOMongoToOneDescription new attributeName: 'planet'; accessor: (MAPluggableAccessor read: [ :paper | paper planet ] write: [ :paper :planet | paper planet: planet ]); yourself! ! !Paper class methodsFor: 'as yet unclassified' stamp: 'sabineknoefel 12/2/2013 12:23'! mongoStar <mongoDescription> ^ VOMongoToOneDescription new attributeName: 'star'; accessor: (MAPluggableAccessor read: [ :paper | paper star ] write: [ :paper :star | paper star: star ]); yourself! ! -- View this message in context: http://forum.world.st/Mongo-lost-references-bug-tp4726673.html Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

3 4