Squeak-dev May 2005

squeak-dev@lists.squeakfoundation.org

153 participants
243 discussions

Congratulation to SM account #300!
by goran.krampe＠bluefish.se 26 May '05

26 May '05

Just wanted to say welcome to Jim Buckeye - SqueakMap account holder number 300! :) regards, Göran PS. Btw, people who knowingly have created SM accounts under "phony" names - I will later push to clean those up, just so you know. I want SM accounts to have proper real names with first and last name.

1 0

Early Registration
by Jason Jones 26 May '05

26 May '05

All, Today is the last day for early registration for Smalltalk Solutions. After today, all registration fees will increase by $100. Please visit the Smalltalk Solutions web site at <http://www.smalltalksolutions.com/registration2005.htm> http://www.smalltalksolutions.com/registration2005.htm to register. Jason Jones Knowledge Systems Corporation Smalltalk Industry Council 919.789.8549 x.21 www.ksc.com www.stic.org www.smalltalksolutions.com

1 0

Connecting a socket to a server.
by François THIMON 26 May '05

26 May '05

Hi. I'm trying to use a client socket that connects itself to a server (from KomServices) that's already running in squeak. The socket seems to be created correctly but the connection primitive always fails. I've tried the following statements : Socket newTCP connectTo:'localhost' port:1024 Socket newTCP connectTo:'127.0.0.1' port:1024 Socket createIfFail:[...] connectTo:'localhost' port:1024 Socket createIfFail:[...] connectTo:'127.0.0.1' port:1024 Does someone know why? Thanks in advance.

2 2

Re: [Q] About compressing
by Lex Spoon 26 May '05

26 May '05

Edgar posted some ZIP-ing code and asked about gzipping. It turns out that there are some gzip examples in the image. (1) The "mail to list" feature in the change sorter uses gzipped change sets. If you know this (and now you do :)), you can do the following to find the relevant code: 1a. open a change sorter. 1b. bring up the menu in the top-left pane. 1c. get a halo around the "mail to list" menu item 1d. explore it, seeing that #mailOut is the relevant selector 1e. Look for implementors-of mailOut (2) You can also simply look at users of class GZippThingamabob. Open a browser, find-class "gzip", and then "class refs" on the resulting class. There are some good-looking examples, including the "mail to list" one and one that compresses an arbitrary file. I hope this helps. These kind of code-navigation skills are very helpful for using Squeak; a lot of the "bad" documentation (not all!) of Squeak is because the old hats are good at these things and quietly assume that you will be as well. On your other question, Edgar, about compressed sources, I don't know the details but there is an intriguing method named #compressSources.... Lex

2 3

Re: Bug in nested iterations? (was: What is wrong with this?)
by Joaquin Sitte 26 May '05

26 May '05

Hi Yoshiki, >> map_ Som new: 30 >> map from: 1 to: map size put: (FloatArray new: 4). > This sets it up so that only one FloatArray instance is created and >it will be pointed to by all the 30 slots of 'map' object. > (map first identityHash) = (map last identityHash) >will be true. This explains the behaviour. But this does not seem to be a terribly useful way of doing things, at least not for my purpose. What is the proper way to define a separate vector for every element in the map? Many thanks Joaquin ===================================================== Dr. Joaquin Sitte, Associate Professor. Head, Smart Devices Lab School of Software Engineering and Data Communication Faculty of Information Technology Queensland University of Technology GPO Box 2434, Brisbane, Q 4001 Australia Phone +61 7 3864 9325 Fax +61 7 3864 9390 e-mail j.sitte(a)qut.edu.au homepage <http://www.fit.qut.edu.au/~sitte> http://www.fit.qut.edu.au/~sitte

3 3

Squeak, Seaside and SqueakMap
by Benoit St-Jean 26 May '05

26 May '05

Hi everyone I have been trying to experiment with Seaside but without any success... I've tried to get the latest packages with SqueakMap package loader with a 3.7 and 3.8 image but I always get walkbacks while loading... What would you recommend as a stable setup to be able to load KomHttpServer and Seaside ? Which image? Thank you. ----------------- Benoit St-Jean MySQL DBA Yahoo! Messenger: bstjean

4 3

Re: Bug in nested iterations? (was: What is wrong with this?)
by Joaquin Sitte 26 May '05

26 May '05

Hi Hans-Martin, Thank you for your reply. I had already checked out the random number generation and concluded that the problem is not with the random number generator. To see this try the following: randomGenerator _ Random new. randNumbers _ FloatArray new: 10. 1 to: randNumbers size do: [:each | randNumbers at: each put: randomGenerator next]. This gives 10 different random numbers. By repeating the last message a new set of random numbers is produced (next takes care of the seeding), as can be verified by inspecting randNumbers. Regardless of the random numbers, the problem I have is the "strange" behaviour of nested iteration. I expect nested iterations in Squeak to behave like nested do loops in procedural programming languages. The "strange" behaviour is evidenced by the following example that does not use the random number generator. Define the class Som ArrayedCollection variableSubclass: #Som instanceVariableNames: 'width height' classVariableNames: '' poolDictionaries: '' category: 'NeuralNets' map_ Som new: 30 map from: 1 to: map size put: (FloatArray new: 4). count _0. map do: [:each | 1 to: each size do: [:i | each at: i put: (count _ count +1)]]. I would expect that the elements of the FloatArray vectors have values starting at 1 and running through to 120. Not so! All elements of the Som end up with the same vector, namely #(117.0 118.0 119.0 120.0) Is this a bug in Squeak or am I missing something? Thanks for any help. Joaquin ===================================================== Dr. Joaquin Sitte, Associate Professor. Head, Smart Devices Lab School of Software Engineering and Data Communication Faculty of Information Technology Queensland University of Technology GPO Box 2434, Brisbane, Q 4001 Australia Phone +61 7 3864 9325 Fax +61 7 3864 9390 e-mail j.sitte(a)qut.edu.au homepage http://www.fit.qut.edu.au/~sitte ------------------------------ Message: 18 Date: Wed, 25 May 2005 12:13:52 +1000 From: "Joaquin Sitte" <j.sitte(a)qut.edu.au> Subject: What is wrong with this? (Beginner question) To: <squeak-dev(a)lists.squeakfoundation.org> Message-ID: <200505250214.CWE38339(a)mail-router02-eth0.qut.edu.au> Content-Type: text/plain; charset="us-ascii" In the course of implementing a Self-organising feature map (SOM) I needed to initialise the weight vecorts of each of nodes in the map with random numbers. I wrote the method randomWeights below for the class Som ArrayedCollection variableSubclass: #Som instanceVariableNames: 'width height' classVariableNames: '' poolDictionaries: '' category: 'NeuralNets' randomWeights "initialise the weights of the neurones to random vectors" "JS 5/24/2005 23:14 Caution: all rendome vectors are the same. Why?" |randomGenerator | randomGenerator _ Random new. self do: [ :each | 1 to: each size do: [:index | each at: index put: randomGenerator next]]. It turns out that the weight vectors of each of the nodes in the SOM are set to the SAME random vector, when I expected them to be all DIFFERENT. What is wrong with this nested iteration? The following trial in a workspace map _ Som new: 30 map width:6 height:5 inputs: 4 count _1. 1 to: 30 do: [ :each | (map at: each) at: 1 put: count. Count _ count + 1]. sets the first weight of each of the 30 nodes to 30 !!! I would expect them to be set to 1, 2, ... 30. Please help. Thanks Joaquin Thanks for your comments Hans-Martin. The random number generator is initialised outside the iteration. Try out the following: myVector _ FloatArray new: 10 1 to: myVector size do: [:each | myVector at: each put: randomGenerator next] ===================================================== Dr. Joaquin Sitte, Associate Professor. Head, Smart Devices Lab School of Software Engineering and Data Communication Faculty of Information Technology Queensland University of Technology GPO Box 2434, Brisbane, Q 4001 Australia Phone +61 7 3864 9325 Fax +61 7 3864 9390 e-mail j.sitte(a)qut.edu.au homepage http://www.fit.qut.edu.au/~sitte ------------------------------ Message: 20 Date: Wed, 25 May 2005 07:18:29 +0200 From: Hans-Martin Mosner <hmm(a)heeg.de> Subject: Re: What is wrong with this? (Beginner question) To: The general-purpose Squeak developers list <squeak-dev(a)lists.squeakfoundation.org> Message-ID: <42940AA5.4090504(a)heeg.de> Content-Type: text/plain; charset=ISO-8859-1 Joaquin Sitte wrote: > It turns out that the weight vectors of each of the nodes in the SOM are set > to the SAME random vector, when I expected them to be all DIFFERENT. What is > wrong with this nested iteration? You have most likely initialized all your Som instances with the same vectors. Since your example does not show the initialization, I can't show you where exactly. Cheers, Hans-Martin

3 2

submorphs overlapping
by François THIMON 26 May '05

26 May '05

Hi, I'd like to now how it was possible to draw the submorphs UNDER their owner instead of drawing them ABOVE it. Thanks

3 2

Camp Smalltalk at StS 2005
by Jason Jones 25 May '05

25 May '05

There will be tables and such set up in the exhibit hall that can be used for Camp Smalltalk. Internet access will be available as well in the conference area. Beforehand we don't, unfortunately have access to space. There is also internet access in the rooms, free if you join Wydham by request. If people are there early there could be some informal Camp activity run out of people's rooms, or in the lobby. Remember that early registration ends May 26 so sign up today and save money on the show. Visit www.smalltalksolutions.com to register. -- Jason Jones Knowledge Systems Corporation Smalltalk Industry Council 919.789.8549 x.21 www.ksc.com www.stic.org www.smalltalksolutions.com

1 0

Linux High Util issues.
by Gerard Korsten 25 May '05

25 May '05

Hi all, I'm currently running the VM: Squeak3.6 of '6 October 2003' [latest update: #5429], , with any of the images (3.6, 3.7 and 3.8) and I'm having some issues with High Utilization on a Gentoo Linux box Whenever I open a MVC project, squeak hogs my cpu at 96% on a constant basis. This is a new empty MVC project, that contains nothing. It also opens slowly, and opening windows (like a browser) takes a bit of time. On a Morphic project everything is fine, and my squeak seldom uses 5% of the cpu. Does anybody else have similar issues, and perhaps any idea where I can start looking to see what in the squeak mvc project is hogging the cpu ? Thanks Gerard

2 2

← Newer
1
...
4
5
6
7
8
9
10
...
25
Older →

Jump to page: