Magma February 2006

magma@lists.squeakfoundation.org

9 participants
9 discussions

Double the size
by Brian Murphy-Dye 01 Mar '06

01 Mar '06

The following code snippet shows a size of 10, but #do: processes those 10 items twice. Any ideas what I'm doing wrong? m := MagmaCollection new. 1 to: 10 do: [:each | m add: each]. Transcript show: String cr, 'size: ', m size asString, String cr. m do: [:each | Transcript show: each asString, String cr]. Brian.

3 3

[ANN] Latest Lava
by Brent Pinkney 26 Feb '06

26 Feb '06

Hi all, The latest version of 'Lava' and 'Lava tester' is available on Squeaksource. This version fixes a few failing unit tests. For those new to Magma, Lava is a package which allows a Magma repository to appear to be a relational database. The developer defines a relational schema, by mapping collections to tables and attributes to columns. SQL queries can be executed against the tables in the schema; these are translated to Smalltalk code and executed against the Magma repository. Now, if only someone would volunteer to write an ODBC driver... PS. This requires Magma1.0-cmm.4.mcz All questions welcomed. Cheers Brent

1 0

Indexing
by goran＠krampe.se 24 Feb '06

24 Feb '06

Howdy folks! Ok, I am using indexes and have a question: My scenario right now (I will probably work around it, but would like to know how this is meant to work) is that I have two different MagmaCollections - one holding n objects and the other holding a subset of those. So the same object occurs in two MagmaCollections. Now, both collections have indexes. Let's say I navigate using one collection to object A, call noteOldKeysFor: on it, then change it and commit. AFAICT MaTransaction>>markRead:using: calls monitorLargeCollectionChanges:, but the WeakSet in MaTransaction just keeps growing, I assume it uses identity instead but perhaps should use equality?. Looks like a bug, anyway that is not my issue here... So the code in captureOldHashesFor: seems to only capture the hashes for the large collections that are "monitored", but in my case this happens to only include the collection I navigated through (since I then read it), but not the other! Is this logic really correct? Or do I need to "trick" the session into reindexing my other collection too somehow? The current result seems to be that object A is only reindexed in the collection I navigated through. I assume there is currently no way for Magma to know which collections it should re-index on its own. regards, Göran

2 4

Magma1.0-cmm.2.mcz
by Chris Muller 23 Feb '06

23 Feb '06

I just posted some fixes to 1.0. I'm going to hold off on posting a new 1.1. I'd like to reiterate that 1.1 is still "in a state of flux" as I review and improve the security. Feel free to look at it for an overall idea of what it will be, but I would hold off migrating to it because a new one will be posted pretty soon. Thanks, Chris - refreshPersistentObjectsEvenWhenChangedOnlyByMe now works consistently with MagmaCollections. When false, all large-collection changes are kept through a begin or abort. There now is no public API for "canceling" large-collection changes other than setting refreshPersistentObjectsEvenWhenChangedOnlyByMe: true, and then aborting. - MutatingProxy fix. - Some how lost this change introduced in version 44. I restored it: "- To support a significant performance improvement, all MagmaCollectionIndexes must have their recordSize set to a power of two. Thanks Andrew Gaylard!"

1 0

Re: [Seaside] (no subject)
by Chris Muller 21 Feb '06

21 Feb '06

Hi Rich, this is a reply to your Magma question posted on the Seaside list so as not to intrude on the Seaside discussions. Magma is supported, so please feel free to ask any questions you have here, I or someone will gladly help you. > So I've looked at Magma. The problem is, it looks > like Magma has had > a few recent changes, and seems to be having > problems. Yes, there are recent changes and more are planned soon. What sort of problems are you having? > I'm having > trouble finding up-to-date documentation. All of the documentation is available on the Swiki and has been updated within the last month. http://minnow.cc.gatech.edu/squeak/2665 Is there some page somewhere that has an older code-example? A reference would be most helpful. > I THINK I > finally have it > set up properly (though there is a strange enter > password to unlock > keychain/enter password to save keychain two-step > dance I have to go > through every time I start up the squeak image, > which just feels > wrong). Ok, you have downloaded the latest 1.1 version which includes the first security implementation. The "two-step dance" you refer to is required to ensure the security of your database. You are supposed to insert your keychain-microdrive to get access to the db. If you don't want security then you should turn it off or use the 1.0 version which does not include security. The swiki has up-to-date documentation for how to turn it off. Please see: http://minnow.cc.gatech.edu/squeak/5832 under the heading "Muting Security". > More importantly, the MagmaSession method > connectAs: seems to > have vanished (though the magma seaside stuff (e.g. > WAMagmaSession) > still try to use it). Ah, ok, the Magma Seaside demo is still based on 1.0. In 1.0 there is no security, so the demo is able to use #connectAs: aString. In 1.1, a MagmaUser with various credentials is supplied automatically. I have updated the instructions for the "Magma and Seaside" page to load "Magma1.0" instead of the newer 1.1 until we update the demo. Sorry about that confusion. > I've tried to fix it by using > connect: (which > did not work) and connect (no arguments). The second > attempt worked, > sort of. But when the web browser starts a new > session it takes > several seconds for the web page to load--which > feels too long. #connect is the right method. The initial connection is a known to be a bit slow (about 1-2 seconds) but has an easy solution. Just preallocate your next session in a background process so it is ready to go. > And, > when I have multiple browsers/sessions open, they > work fine at first, > but eventually stop seeing each others changes. So > something's not > right. Hard to know what the problem is here.. No one else has reported this. Can you offer more details? > So, does anyone have a better recommendation for a > database that I > can move easily between Mac and PC? Or does anyone > know a trick to > getting the new version of Magma to work? Any > suggestions are welcome. If I may assist you with Magma, please let me know. Regards, Chris

1 0

Running MagmaBenchmarker
by Paulo Sérgio Medeiros 19 Feb '06

19 Feb '06

Well, i know that it can be obvious (since i'm new to all that's surrounds squeak) but i spend some minutes figuring out how to run MagmaBenchmarker (looking throught the code). I didn't find this information on the site ( http://minnow.cc.gatech.edu/squeak/5606): mb := MagmaBenchmarker localUsingWriteBarrier: false rpoewcobm: false. st := mb runAll. st inspect. Cheers, Paulo Sérgio. PS: Also, only for curiosity, here is my result: 'The date is 19 February 2006 7:19:06 pm Hardware Details: computer : pc cpu : AthlonXP (Barton) speed : 2.35GHz memory : 1024MB disk : internal IDE OS Details: osVersion : NT platformName : Win32 platformSubtype : IX86 vmVersion : Squeak3.7 of ''4 September 2004'' [latest update: #5989] imageName : E:\smalltalk\Squeak\3.8.14.image Image Details: version : Squeak3.8 lastUpdate : 6665 Code Package Details (from Monticello): ** modified ** Name: Magma tester-cmm.223 Author: cmm Time: 6 January 2006, 3:53:59 pm UUID: 59f02b20-c95c-3a45-a4df-1f09451bf19d Ancestors: Magma tester-cmm.222 Magma server details: flushAggressively : false MagmaSession Details: isLocal : true allowWriteBarrier : false refreshPersistentObjectsEvenWhenChangedOnlyByMe : false Benchmarker Details: thousands : 1000 Benchmarks: --- connectionTests connect : 0:00:00:01.038 --- readTests peakRefreshRate : 5374.12517496501 per second. singleObjectRead : 3075.584883023395 per second. oneThousandElementArrayRead : 346.9918049170498 per second. oneThousandElementArrayOfObjectsRead : 8.21274931560422 per second. oneMillionObjectPointersRead : 0.299132515704457 per second. oneThousandLevelsDeepRead : 7.29639124433051 per second. --- writeTests peakCommitRate : 14.42687747035573 per second. oneThousandElementArrayCommit : 12.2432859399684 per second. a1001BufferCommit : 9.5068330362448 per second. aOneMillionBufferCommit : 10 seconds. --- magmaArrayTests getMagmaArraySize : 6797.692990239574 per second. updateThousandsOfMagmaArrayElements : 0:00:00:00.008 --- magmaCollectionTests addThousandsOfObjectsTenAtATime : 12.552 seconds. addThousandsOfObjectsOneHundredAtATime : 3.182 seconds. addThousandsOfObjectsAtOnce : 7.83 seconds.'

1 0

Next steps
by goran＠krampe.se 12 Feb '06

12 Feb '06

Hi Chris and all! I am using Magma now - though I haven't stressed it yet at all, still busy building my domain model. But it seems to "just work" for the moment - which is really nice. I noticed: http://minnow.cc.gatech.edu/squeak/5832 ...and accidentally that is exactly what I need. :) Now, it doesn't necessarily fit me - so I am using a prevaylerish approach with "command" objects that I also save to Magma so that I can reconstruct the model by applying the commands in sequence. This should give me a suitable mechanism to keep Magma database copies (offline clients) in synch with the master and also reasonable hooks to figure out "logical" conflicts. But I was just curious if you could share your current "thoughts" on the subject. Also, the security stuff in 1.1 is interesting - we might want to use repository file encryption (or just rely on OS level mechanisms for that - encrypted filesystems). Was there any other relevant changes with 1.1? regards, Göran

4 7

Error in KryptOn/Magma SUnit Tests
by Dane Jensen 07 Feb '06

07 Feb '06

As the first step in getting the hang of Magma, I'm trying to load Magma into a fresh image and run the sunit tests. So far so good with the loading, not so good with the tests though. I'm getting a "MessageNotUnderstood: Rijndael class>>asMakoEncryptable:" in "MakoSecurityManager>>basicEncrypt:using:" error from the Magma tests. I assumed it had something to do with KryptOn and ran the tests for it, same error. I'm probably missing some simple package somewhere, but have no clue which. Any ideas? Thanks! -Dane

3 3

Moving forward
by goran＠krampe.se 06 Feb '06

06 Feb '06

Hi Chris and the rest! Just wanted to report that the demo we held last friday of our "Q2" software prototype was a success and that we are moving forward with the project. As it looks now we are working with a milestone each month into july. Magma works great so far, I will be starting to report on our experiences later on. regards, Göran

3 2

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

Magma February 2006