Magma January 2010

magma@lists.squeakfoundation.org

9 participants
12 discussions

Fwd: [ANN] PDF of Dynamic Web Development with Seaside
by Mariano Martinez Peck 27 Jan '10

27 Jan '10

For those who may be interested. ---------- Forwarded message ---------- From: Lukas Renggli <renggli(a)gmail.com> Date: Wed, Jan 27, 2010 at 2:49 PM Subject: [squeak-dev] [ANN] PDF of Dynamic Web Development with Seaside To: Seaside <seaside(a)lists.squeakfoundation.org>, Pharo Development < pharo-project(a)lists.gforge.inria.fr>, Squeak < squeak-dev(a)lists.squeakfoundation.org>, "Magritte, Pier and Related Tools ..." <smallwiki(a)iam.unibe.ch> The PDF version of the book "Dynamic Web Development with Seaside" is available for download now: http://book.seaside.st/book/introduction/pdf-book At the end of the payment process (PayPal) you will be redirected to the download area where you are able to get the latest builds of the PDF version of the book. If you bookmark the page you will be able to download fixes and extra chapters as we integrate them into the online version. By buying the PDF version you support our hard work on the book. We wish to thank the European Smalltalk User Group (ESUG), inceptive.be, Cincom Smalltalk and GemStone Smalltalk for generously sponsoring this book. We are looking for additional sponsors. If you are interested, please contact us. If you are a publisher and interested in publishing this material, please let us know. Please distribute this message widely. Cheers, Stef & Lukas -- Lukas Renggli http://www.lukas-renggli.ch

1 0

Magma and process
by Juan 25 Jan '10

25 Jan '10

Folks I have a new project using magma, but by testing with a simple model I could see that the process is filled processBrowser in wait. Many of them with names like many with names like 'Time to Load Buffers' 'Time to Download Full Response' 'Time for Initial Response' etc etc.. The number of processes is also striking for example now I'm 82. some background Magma client-cmm.446 Magma seaside-brp.4 Magma server-cmm.382 Magma has a lot of packages and not as I determine which version, if there is a way to load a version stable of monticello please let me know Any idea will be appreciated best mdc

3 2

[Q] Any simple magma example for text/tags combination?
by Sungjin Chun 19 Jan '10

19 Jan '10

Hi, I'd like to test magma with following data; 1. it contains text (string) as main data 2. it contains title (string) 3. it contains tags (a tag is string) I like to acess/order the data by title and want to access data by specifying a tag. Sorry for poor description and thanks in advance.

2 3

how to do Magma Class Evolution?
by Bart Gauquie 18 Jan '10

18 Jan '10

Dear all, I'm trying out the Class Evolution features of Magma. Typically you want to have following stuff working: rename class, rename field, totally change internal structure of some part of the domain. You do this during development and have to make sure the data in production database gets converted to these new changes. I started by testing the 'rename class scenario': You have a running production system with a filled database containing objects of a certain type. During development, a class is renamed. The first thing I tried was to load the package containing the new class type (and not the old type anymore) in the production image. As promised on http://wiki.squeak.org/squeak/5602, under implementation, Magma recreates the old class definition in the production image. However, the objects defined in the database are not automatically converted to the new classes. (the rename of the class happened in the developer image, not in the production image); meaning that if you add new entries, objects of both class types will be in the database (which is bad). To convert the old classes in database to the new classes, I tried the update strategies, as mentioned on http://wiki.squeak.org/squeak/5602. Working out the second option: one-shot script to enumerate all instances and populate, I followed these steps (my domain is about Productbacklogs, sprints and UserStories): - Userstory >> various >> copy: UserStoryOld (make a copy of the old UserStory) - UserStory >> refactor class >> rename: UserStoryNew (give the UserStory the new name) - UserStoryOld >> refactor class >> rename: UserStory (now you have a copy of the class definition of production with the old name) - write a script which converts all entries in the database using UserStory to UserStoryNew UserStory>>asUserStoryNEW |result| result := UserStoryNew new businessvalue: self businessvalue; contents: self contents; title: self title; yourself. resterendeTijdenNodig do: [:each| result addResterendeTijdNodig: each]. ^result and: a script which iterates over all userstories in the database, creates UserStoryNew's from them, makes sure all links from and to the old UserStories are replaced by the new ones (for instance: put them in a productbacklog, replan them, ...); saving each of them again. - Load this package in the production image, run the script => all old entries are replaced with new classes. This approach works. It has the advantage that every change is committed, you can rollback to a commit point if something fails, and so forth. But it takes a lot of time to do such a simple change? Is this what is meant by a 'one-shot script to enumerate all instances and populate' ? Is there another way of doing this change? Because it would mean that for every rename of a class you would have to write such a script. A field rename can be handled in the same way. Let the old and new field coexist, a script iterates over all the objects, copying the old field to the new one. If it is just a simple type (Integer, Boolean or String), this will be simple again, however if this field is something complex with a lot of pointers from and to it, it will become somewhat more difficult to correctly write a convert script. But it will be possible. According to the wiki there is also 'a faster one-shot script that operates directly on the MaObjectBuffers using MagmaFileTraverser'. I looked into this, but I did not understand it. Any pointers how to, for instance, do a class rename using this technique? Thanks for any advice, Kind Regards, Bart -- imagination is more important than knowledge - Albert Einstein Logic will get you from A to B. Imagination will take you everywhere - Albert Einstein Learn from yesterday, live for today, hope for tomorrow. The important thing is not to stop questioning. - Albert Einstein The true sign of intelligence is not knowledge but imagination. - Albert Einstein However beautiful the strategy, you should occasionally look at the results. - Sir Winston Churchill It's not enough that we do our best; sometimes we have to do what's required. - Sir Winston Churchill

2 1

Metacello configuration of Magma
by Miguel Enrique Cobá Martinez 14 Jan '10

14 Jan '10

I have published the Metacello configuration of Magma in the MetacelloRepository of squeaksource.com. Take a PharoCore 10506 and evaluate in a workspace: Gofer it squeaksource: 'MetacelloRepository'; package: 'ConfigurationOfMagma'; load. Then load the default group that loads the Magma Client: ((Smalltalk at: #ConfigurationOfMagma) project version: '1.0r43') load this is equivalent to: ((Smalltalk at: #ConfigurationOfMagma) project version: '1.0r43') load: 'Client'. To load the Magma Server evaluate: ((Smalltalk at: #ConfigurationOfMagma) project version: '1.0r43') load: 'Server'. And to load the Magma Tester evaluate: ((Smalltalk at: #ConfigurationOfMagma) project version: '1.0r43') load: 'Tester'. This last one will load also OSProcess that is used by the Magma Tester to do the stress testing of Magma. Notes: - WriteBarrier is a package loaded by Magma. This requires NewCompiler as dependency but as NewCompiler is being rewritten currently, isn't loaded as a dependency. - Magma makes use of ProjectHistory that doesn't exist anymore in Pharo. Magma will fix this by conditionally testing for ProjectHistory by means of Smalltalk at: #ProjectHistory. - This is the first configuration for Magma and will be modified in response to the feedback you give me. - Currently the blessing is #development. Maybe this should be marked as #stable in order to load with: ((Smalltalk at: #ConfigurationOfMagma) project latestVersion) load. This way it will install the latest non-development version available in MetacelloRepository without having to specify a version number. - This configuration executes a preLoadDoIt with the following code: Smalltalk at: #EventSensor put: InputEventSensor because EventSensor was removed from Pharo, but Magma add some methods to EventSensor class. I *really* need feedback about the risks of doing this. Any feedback is welcome. Enjoy! -- Miguel Cobá http://miguel.leugim.com.mx

2 2

Re: Metacello ConfigurationOfMagma
by Miguel Enrique Cobá Martinez 13 Jan '10

13 Jan '10

The message was held until approvement by attachment size. I have exported png to gif with very bad quality and zip the files. They are attached now Cheers El mar, 12-01-2010 a las 12:08 -0600, Miguel Enrique Cobá Martinez escribió: > Hi Chris, > > I'm creating the Metacello configuration for Magma. I have almost all > done but I need your feedback for the correct dependencies of the > packages that compose Magma. > > First, I used a clean PharoCore 10505 image. > Using monticello browser I loaded > > 1.0r43 mcm -> 1.0r43 (tester).mcm > > that installed: > > Ma exception handling-cmm.32 > Ma base additions-cmm.149 > Ma proxy support-cmm.40 > Ma traverse object graphs-cmm.29 > MaFixedWidthReport-cmm.7 > Ma Statistics-cmm.21 > Ma object serialization-cmm.220 > Collections-BTree-lr.68 > Ma special collections-cmm.101 > Ma client server-cmm.214 > WriteBarrier-pmm.26 > Magma client-cmm.446 > Ma contextual search-cmm.29 > Magma server-cmm.382 > Ma Armored Code-cmm.144 > Ma Squeak domain-cmm.28 > Ma special collections tester-cmm.12 > Ma object serialization tester-cmm.27 > Magma tester-cmm.352 > > Then, using the instructions from Lukas Renggli: > > http://lists.gforge.inria.fr/pipermail/pharo-project/2010-January/019678.ht… > > and using the DependencyBrowser from Hernán Morales: > > http://cs.hernanmorales.com.ar/projects/dependencyBrowser/DBrowser-en.php > > I created the attached .dot file with the dependencies for all the > packages for Magma (tester, server and client). > The code to generate it was: > > (PDPackageAnalyzer onPackagesNamed: #( > 'Magma tester' > 'Ma special collections tester' > 'Ma Squeak domain' > 'Ma Armored Code' > 'Magma server' > 'Ma contextual search' > 'Magma client' > 'WriteBarrier' > 'Ma client server' > 'Ma special collections' > 'Collections-BTree' > 'Ma object serialization' > 'Ma Statistics' > 'MaFixedWidthReport' > 'Ma traverse object graphs' > 'Ma proxy support' > 'Ma base additions' > 'Ma exception handling' > )) save: 'magma.dot' > > Then using graphviz from linux (installed in Debian Linux as aptitude > install graphviz) I generated the attached png graph of the dependencies > with: > > $ cat magma.dot | dot -Tpng > magma.png > > I have checked dependency browser vs. PDPackageAnalyzer and both of them > agree on the dependencies so I am almost sure that is ok. But your > review will be very welcome. > > Couple of things I found in this process: > > - The WriteBarrier depends on class ByteCodeGenerator that isn't > included in Pharo. It appears to be on NewCompiler. Should I add > NewCompiler as a dependency for Magma, or is writebarrier useful without > NewCompiler. > - The Magma tester project depends on ProjectHistory that doesn't exist > anymore in Pharo. Maybe this should be a conditional in Magma or the > test modified to not use this class, at least in Pharo. > - Ma Armored Code depends on OSProcess but the mcm doesn't includes it. > > Now, as I said, I am doing the Metacello configuration for loading Magma > in Pharo (and maybe also in Squeak) and this will permit to load Tester, > Server and Client, WITH ALL the dependencies needed. This means that > when loading tester will load OSProcess and when loading WriteBarrier > will load also NewCompiler. > > Any comment on this? > > Thanks > > -- Miguel Cobá http://miguel.leugim.com.mx

3 10

Dependencies of Magma 1.0r43
by Miguel Enrique Cobá Martinez 13 Jan '10

13 Jan '10

This is the list I compiled for the Magma Dependencies in the creation of the Magma Metacello configuration: WriteBarrier -> none (BytecodeGenerator) Collections-BTree -> none Ma exception handling -> none Ma base additions -> Ma exception handling Ma proxy support -> Ma exception handling Ma Squeak domain -> Ma base additions Ma contextual search -> Ma base additions MaFixedWidthReport -> Ma base additions Ma special collections -> Ma base additions, Collections-BTree Ma traverse object graphs -> Ma proxy suport Ma Statistics -> MaFixedWidthReport Ma object serialization -> Ma Statistics, Ma special collections, Ma traverse object graphs Ma client server -> Ma object serialization Ma Armored Code -> Ma client server, [OSProcess] Magma client -> Ma client server, WriteBarrier Ma object serialization tester -> Ma Squeak domain, Ma contextual search, Ma Armored Code Magma server -> Ma contextual search, Magma client Ma special collections tester -> Ma Armored Code, Magma client Magma tester -> (ProjectHistory), Ma object serialization tester, Magma server - BytecodeGenerator isn't loaded by default in Squeak or Pharo. Is part of NewCompiler but currently is being rewritten for the closure image. - ProjectHistory only exists in Squeak. In Pharo was removed long ago because of the unloading of the Project code. - OSProcess is used by the tester but isn't loaded by default in Squeak nor Pharo. I attach the updated .dot file and jpg file of the dependencies graph of Magma 1.0r43 (tester).mcm. Cheers -- Miguel Cobá http://miguel.leugim.com.mx

2 4

Re: Querying nested MagmaCollections
by Amir Ansari 08 Jan '10

08 Jan '10

Thanks so much! That's an ingenious solution! Yes, there were a couple of 'bugs' ;-) but the concept was the most important thing. I have it working perfectly now. Regarding alternative solutions: one could also make separate 'tables' for books and writers (I didn't use 'author' since it's already an important class in the Squeak system), related to each other by a primary key. (The classic relational model.) But I think it's inelegant for an object oriented system, which is why I was searching for a neater solution. I'll have a look at Lava too! Amir

1 0

Why is MagmaSession>>#noteOldKeysFor: necessary?
by Bart Gauquie 08 Jan '10

08 Jan '10

Dear all, I'm using Magma and finding out how transparent it can be. I'm trying Magma to see if the DDD principles can be applied to it. To retrieve, store and save aggegrate root objects a Repository object is used. A 'transient' Repository object is nothing more than a wrapper around an OrderedCollection, which contains store save and find methods. The client who is using this Repository object does not know which implementation is behind it. After retrieving the initial root aggegrate the client can manipulate the sub objects of the aggegrate root. I managed to achieve Read Transparency. I specialised the Repository objects so they use a MagmaCollection instead of an ordinary OrderedCollection. You just do a findXXX method on the Repository, and that method is delegated to a MagmaCollection which does a read based on some defined index. To do a Write Transparency, I added a decorator around the store / save methods of the Repository which automatically wraps the original Repository object in a session commit:[] block; enabling Magma to see the changes made to the 'saved' object and get them persisted. AddTransactionAroundMutators>>doesNotUnderstand: aMessage |result| (self isMutator: aMessage) ifTrue: [repositoriesController commit: [result := aMessage sendTo: decorated]] ifFalse: [result := aMessage sendTo: decorated]. ^result. In which AddTransactionAroundMutators is a decorator around the Repository object. Persisting works, however I noticed that the hash indexes did not get updated this way. On the wiki it is mentioned that MagmaSession>>#noteOldKeysFor: should be called before any change to the object is made. This is off course, using my idiom, impossible to do, since the user using the Repository does not even know that there is a MagmaSession. It just knows the API of the Repository object. The Seaside GUI I'm using is manipulating an object (which it got from a findXXX query on a repository) directly and if it wants to save the changes it calls Repository>>save (which got decorated). The GUI using the Repositories does not know Magma and I tend to keep it that way. After some searching, I managed to circumvent this by a dirty hack: AddTransactionAroundMutators>>doesNotUnderstand: aMessage |result| (self isMutator: aMessage) ifTrue: [ (self isUpdatingMutator: aMessage) "denotes saveXXX methods => updates to existing objects" ifTrue: [ |session session2 firstArg oidForFirstArg reReadObject| session := repositoriesController session. session2 := repositoriesController session2. firstArg := aMessage arguments first. oidForFirstArg := session oidFor: firstArg. reReadObject := session2 realObjectFor: oidForFirstArg. repositoriesController commit: [session noteOldKeysFor: firstArg usingHashesFromObject: reReadObject. result := aMessage sendTo: decorated]] ifFalse: [ repositoriesController commit: [result := aMessage sendTo: decorated]]] ifFalse: [result := aMessage sendTo: decorated]. ^result. MagmaSession>>noteOldKeysFor: anObject usingHashesFromObject: anOtherObject guard critical: [ transaction ensureOldHashesCapturedFor: anObject usingHashesFromObject: anOtherObject] MaTransaction>>ensureOldHashesCapturedFor: anObject usingHashesFromObject: anOtherObject (potentialKeysChange includesKey: anObject) ifTrue: [ ^ self ]. self captureOldHashesFor: anObject usingHashesFromObject: anOtherObject MaTransaction>>captureOldHashesFor: anObject usingHashesFromObject: anOtherObject largeCollectionChanges do: [ :eachChanges | (eachChanges collection canIndex: anObject) ifTrue: [ eachChanges collection indexesDo: [ :eachIndex | | oldKeys oldHashes | oldKeys := potentialKeysChange at: anObject ifAbsentPut: [ IdentityDictionary new ]. oldHashes := oldKeys at: eachIndex attribute ifAbsentPut: [ OrderedCollection new ]. oldHashes add: { eachChanges. eachIndex. eachIndex indexHashesFor: anOtherObject } ] ] ]. So I'm using a second MagmaSession to reload the object I'm about to commit and I'm noting to the first session to noteOldKeysFor: anObject usingHashingFromObject: theJustReloadedObjectFromTheSecondSession. Like this, the first session gets the hashes of the unmodified object in the repository, and the hashes get updated correctly automatically. It is a working solution (at least with my very limited testing), but I find it very complicated. Surely there must be a cleaner way to do an automatic hashkey update. Other things I already tried is to enable WriteBarrier, but could not get that working. I also tried to do readRealObject: oid on the same session; but this always returned a cached version, with the already updated fields in. The only thing that worked was to create a new session and make sure that the object is fresh. Can anyone help me on this one? I'm finding it a bit strange that Magma can correctly save my changes, but not determine if a hash key update is necessary. Thanks for any help. Kind Regards, Bart -- imagination is more important than knowledge - Albert Einstein Logic will get you from A to B. Imagination will take you everywhere - Albert Einstein Learn from yesterday, live for today, hope for tomorrow. The important thing is not to stop questioning. - Albert Einstein The true sign of intelligence is not knowledge but imagination. - Albert Einstein However beautiful the strategy, you should occasionally look at the results. - Sir Winston Churchill It's not enough that we do our best; sometimes we have to do what's required. - Sir Winston Churchill

2 5

[ANN] Magma version 1.0, release 44, beta2
by Chris Muller 08 Jan '10

08 Jan '10

I have just posted a new version of Magma to SqueakMap, as well as the "MagmaTester" project of squeaksource. It was pretty lame of me to discourage anyone from using a part of Magma that is there to be used. I am therefore pleased, at last, to put to rest this issue with Magma "luxury" queries once and for all. Prior to release 44, beta2, use of the luxury features of a MagmaCollectionReader: - use of a disjunction (|) - requiring distinct results - requiring sort by an attribute other than the optimized attribute would mean the entire query would be executed on the server, even if the application only needed the initial results of the query. This is wasteful, and could cause the server to get bogged down for an application that tried to use frequent, ad-hoc luxury queries against large collections. Now, however, the MagmaCollectionReader>>#release method causes the server resources associated with the luxury query to be cleaned up. This will allow applications to now use these luxury query features for ad-hoc querying. It has *always been* the Magma application-developer's responsibility to send #release to a MagmaCollectionReader when finished with it so, theoretically, there should be no need for any application changes. Before 44Beta2, calling #release had a minimal benefit, merely releasing an in-memory Listener dependent. Now, #release also cleans up significant server resources in the case of a luxury query, a great benefit. If you need scalability, don't let this slide! My hope is that those folks coming to Magma from a relational-db background will appreciate this feature. I do recall, when using Java / Oracle some years ago, that I had to release my ResultSet object when done with it. Therefore, the lack-of-transparency regarding this particular operation will, hopefully, be understood. Despite the improvement, I still encourage developers to look for more OO solutions to app problems. Magma is an ODBMS and works well with large models that do not contain MagmaCollections. MagmaCollections were never intended as the primary domain mechanism of Magma applications. It was always intended as a secondary, supplemental structure (originally intended to solve the use-case of "finding objects by keyword"). It has been tested in Squeak 3.9 and Pharo, but if anyone has any questions or issues, please let me know. - Chris

1 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

Magma January 2010