Squeak-dev January 2006

squeak-dev@lists.squeakfoundation.org

164 participants
300 discussions

Anyone know the status of the Graphs package listed on SM?
by Peter Crowther 02 Jan '06

02 Jan '06

http://kilana.unibe.ch:8888/Graph/Collections-Graphs-sam.19.mcz appears to be a 3.7 release. It depends on two other packages - one for iterators, one for some miscellaneous changes. I'm trying to load it into 3.9a, and am getting an error that KeyedSet cannot be recompiled. Has anyone persuaded this into a later version than 3.7? And how large a hammer did you need? I'm looking at a job that requires some reasonably heavy-duty graph hacking in Squeak, and it'd be good to work with an existing package rather than roll my own. Besides, I trust Mario Wolczko's code more than anything I'd ever write :-). - Peter

2 1

Running sets of unit tests
by Martin Kuball 02 Jan '06

02 Jan '06

Hi I'm looking for an easy wayto run a set of unit tests. The SUnit Test Runner does not allow me to do this. I can only filter by class name which is not sufficient. What I would like to do is to filter by categorie. And running all tests is not an option because it takes way to long to finish. Martin

3 2

[BUG] [FIX] Incremental programming in the debugger
by Diego Gomez Deck 02 Jan '06

02 Jan '06

from preamble: "Change Set: DebuggerFix-dgd Date: 18 December 2005 Author: Diego Gomez Deck <DiegoGomezDeck(a)consultar.com> Fix an error in the incremental programming facilities providen by the Debugger. Error: When a method is implemented using the [Create] button from the debugger, the method is called with only one parameter: The failed message Fix: Moves the arguments from the message to the context temps. " -- Diego -- ============================================================== Diego Gomez Deck ------------------------------------------------------------ http://www.small-land.org http://www.consultar.com/DiegoGomezDeck/ http://diegogomezdeck.blogspot.com/ http://es.groups.yahoo.com/group/LibroProgramacionSmalltalk/ ==============================================================

7 8

Re: Re: [ANN] SmallLint Failure Model
by Hernan Wilkinson 02 Jan '06

02 Jan '06

>Hernan Wilkinson wrote: > >> I uploaded to Squeak Map a package to deal with the SmallLint >> failures. The package name is: Small Lint Failure >> <http://map.squeak.org/accountbyid/ce6957eb-c5ff-4a6e-a575-a1467238905e/pack…> >> The idea of this module is to declare smalllint failures as >> expected, so when SmallLint is run again, the expected failures will not >> appear. > > > >While I'm not familiar with the Squeak version, Smalllint does have the >ability to ignore failures. The BasicLintRule class has a class variable >FilterDictionary that contains a dictionary mapping the lint rule name >to a BrowserEnvironment of methods/classes to ignore. In the VW, VA, & >Dolphin versions, there are filter menu options that display when you >are viewing the lint rule results. There are also options to save/load >the filters to files. I'm guessing that Squeak is just missing this menu >options that expose the functionality. > > >John Brant Hi John, you are right. SmallLint allows you to filter rules but the solution it provides did not fulfill our needs. Although it may work in many situations, we wanted something more "automatic" and controllable. One of the problems we found with the SmallLint filters is that it is difficult to know the filters that you created once they are created, there is not way to "annotate" the reason of the filter or who created it, and the most important for us, was that it did not cover "code evolution". For example, one filter might apply today but if you change the code in such a way that the filter does not apply anymore, you would not know it. The solution we created allows you to write the "expected failures" in such a way that they are part of the class (we use VisualAge with Envy, so it works great when loading different versions of the class), it force you to put a reason and who created it, and it tells you about the "expected failures" that do not apply anymore. Anyway, we use use SmallLint a lot and we thank you for that!. We created this tool and thought that somebody else could found it useful. Bye, Hernan. -- ______________________________ Lic. Hernán A. Wilkinson Gerente de Desarrollo y Tecnología Mercap S.R.L. Tacuari 202 - 7mo Piso - Tel: 54-11-4878-1118 Buenos Aires - Argentina http://www.mercapsoftware.com --------------------------------------------------------------------- Este mensaje es confidencial. Puede contener informacion amparada por el secreto profesional. Si usted ha recibido este e-mail por error, por favor comuniquenoslo inmediatamente via e-mail y tenga la amabilidad de eliminarlo de su sistema; no debera copiar el mensaje ni divulgar su contenido a ninguna persona. Muchas gracias. This message is confidential. It may also contain information that is privileged or otherwise legally exempt from disclosure. If you have received it by mistake please let us know by e-mail immediately and delete it from your system; you should also not copy the message nor disclose its contents to anyone. Thanks. ---------------------------------------------------------------------

1 0

RE: PortAudio
by Cowdery, Bob [UK] 02 Jan '06

02 Jan '06

Brad wrote: > I have been working on it, on and off. It's not as far along as I'd like > it to be. > I've been able to: > set up the plugin to make it and test it (the hardest so far!) > conduct a couple of tests -- calling portaudio individual tests: (play > a sine wave, record a buffer and play back that recording) > I was in the middle of creating individual portaudio methods and finding > out the best way to split responsibilities between squeak and the plugin > when I as interrupted by other (bill-paying) projects. All in Linux, BTW. > Answer to your question: nothing usable. (Unless you want to help!) Thanks for the update Brad. Real work often gets in the way unfortunately. I have a huge work schedule this year mostly on Java, Yuk. I would love to have PortAudio working on Linux and Windows for my Software Radio project. I would also like to help. It would be silly of me to promise any particular level of help and I have to say my ST is literally 2 months or so and I could be so buried the next few months I won't know what day it is. First off if you tell me the magnitude of the remaining work, and the sort of skills I would have to gain. Even if I can only get an environment up, understand what you have done and do some testing I guess that would be a start. Bob *** Confidentiality Notice *** Proprietary/Confidential Information belonging to CGI Group Inc. and its affiliates may be contained in this message. If you are not a recipient indicated or intended in this message (or responsible for delivery of this message to such person), or you think for any reason that this message may have been addressed to you in error, you may not use or copy or deliver this message to anyone else. In such case, you should destroy this message and are asked to notify the sender by reply email.

1 0

RE: Named pipes
by Cowdery, Bob [UK] 01 Jan '06

01 Jan '06

> Load OSProcess from Squeak Map. > Make a named pipe called 'FIFO'. > In Squeak, open the named pipe as a file stream: > > namedPipe := FileStream fileNamed: 'FIFO'. > > Set nonblocking: > > OSProcess accessor setNonBlocking: namedPipe fileID. > > Read from the pipe without blocking (answers an empty string): > > namedPipe next: 10000 > > As you have noticed, a blocking read will hang your Squeak VM if no > data is available. But note that on Linux you can break things loose > by going to /proc/<squeakPID>/fd/ and writing to whatever file > descriptor is waiting for the data. This is handy if you have locked > up Squeak and don't want to lose your work. > > If you set the file descriptor non-blocking with #setNonBlocking:, it > will not hang up the VM. A read will give you the available data > (possible an empty string if no data is ready). > See also AioPlugin on Squeak Map for buffered event-driven input from > a non-blocking file stream. Thanks very much Dave. I figured most of it. The bit that got me was the IOHandle that setNonBlocking wanted was a fileID. I am actually reading audio sampled data in binary mode and get a nil when the pipe is empty. Bob *** Confidentiality Notice *** Proprietary/Confidential Information belonging to CGI Group Inc. and its affiliates may be contained in this message. If you are not a recipient indicated or intended in this message (or responsible for delivery of this message to such person), or you think for any reason that this message may have been addressed to you in error, you may not use or copy or deliver this message to anyone else. In such case, you should destroy this message and are asked to notify the sender by reply email.

2 1

PortAudio
by Bob.Cowdery＠CGI-Europe.com 01 Jan '06

01 Jan '06

I know someone was working on a port of PortAudio to Squeak. Just wondered how far its got and if there is anything usable yet. Thanks Bob *** Confidentiality Notice *** Proprietary/Confidential Information belonging to CGI Group Inc. and its affiliates may be contained in this message. If you are not a recipient indicated or intended in this message (or responsible for delivery of this message to such person), or you think for any reason that this message may have been addressed to you in error, you may not use or copy or deliver this message to anyone else. In such case, you should destroy this message and are asked to notify the sender by reply email.

2 1

Help requested with morphic-related bug
by Kurt Thams 01 Jan '06

01 Jan '06

I wrote a program to solve Sudoku puzzles, and when I run it I get a problem... various submorphs stop drawing (and get rendered as a red box with a yellow cross through it), and a walkback comes up with "Error: Error: Instances of UndefinedObject are not indexable" (in WorldState>>displayWorldSafely) Not sure if I'm doing something wrong with my program or if it is an error in Morphic. Essentially, I fork a process and set a morph to step while the forked process updates the model that the morph is drawing from. The code can be downloaded from here: http://thams.com/Kurt-Sudoku-ktt.2.mcz and executed with 'Kudoku example1' Any help appreciated. -- kurt

3 4

Re: Traits?
by Adrian Lienhard 01 Jan '06

01 Jan '06

Hi all, We have been working on the integration of the traits implementation with current 3.9a. Now it's ready to be tested, played with, reviewed etc. There's a ready image at http://www.adrian-lienhard.ch/files/ Squeak3.9a-traits.zip. For those who like to create this image themselves, here the instructions (of course, this will later automatically work via the update stream): - get http://ftp.squeak.org/3.9alpha/Squeak3.9a-6705.zip (that's the latest published 3.9a image) - open Monticello, select the source.squeakfoundation.org/39a repository - load from there ScriptLoader-al.55 - evaluate "ScriptLoader new updateFrom6705" (which will initiate the traits bootstrap) - Click "proceed" when asked "You are about to load new versions...", same for "Behavior should not be redefined..." - After this, the re-compilation of the image is done, which takes quite some time... and that's it. Note, there's no fancy traits browser as you might expect. What this release is all about is the new class kernel enhanced to support traits, and Monticello and tools have being adapted to properly work with traits. The code browser has a new menu items to display the definition template in order to create a new trait. In addition the kernel includes an algorithm to efficiently calculate the required methods of a class; this was implemented by Nathanael and integrated by Daniel. Daniel has also been working on new tools based on OmniBrowser. This can be installed by loading the latest version of the package from http://kilana.unibe.ch:8888/TraitsOmniBrowser (a prerequisite is latest OmniBrowser from SM). Open the browsers by "OBTraitStructureBrowser new open" and "OBTraitStructureBrowser new open". However, I think, those tools need more work to be ready, and is not subject to inclusion for now. Here an overview for the people interested in looking at the code: - There is a new Kernel package; almost all changes there are in the category Kernel-Classes (Behavior, ClassDescription, ...) - There is a new package named Traits. It includes all additional classes and traits for the kernel (see Traits-Kernel) and also for the requires algorithm mentioned above - This package has a couple of extensions (tip: to look at them, use MC's snapshot browser): most of those class extensions or overrides should eventually be moved to the appropriate packages. - There is a new version of the Tools package which contains the adapted code browser - Quite some changes have been made to Monticello to support saving and loading of traits - Tests can be found in category Traits-Tests I encourage interested people to have a look soon and to bring up any questions or concerns, so that we can avoid stalling the 3.9a process. At this place, I'd like to thank all the people that have been involved in this project in one or the other way. Special thanks to Andrew, Daniel, Marcus, Nathanael and Stef. Cheers, Adrian ___________________ Adrian Lienhard www.adrian-lienhard.ch On Dec 3, 2005, at 21:28 , Cees De Groot wrote: > Hi, > > Any news on traits? Anything else that we should be attending > w.r.t. v3.9? > > Regards, > > Cees

6 10

Squeak VM 1.31 for Windows sources
by Robert Kovacic 01 Jan '06

01 Jan '06

Are the complete source files for the Windows version of the VM for Squeak 1.31 available for download anywhere? Regards, Robert Kovacic

1 0

Jump to page: