Squeak-dev April 1999

squeak-dev@lists.squeakfoundation.org

130 participants
319 discussions

Re: [REQ]Remove InputSensor>>mousePoint
by Reinier van Loon 06 Apr '99

06 Apr '99

Duane, you wrote: >Hmm... not sure about that. It is possible that someone might want the >cursorPoint and the mousePoint to be different things. If we interpret the >mousePoint to be the position of the physical mouse, I can't see why one wants to know the position of the mouse on the mouse pad. > and the cursorPoint to be the position of the indicator on the screen, >then gridding is an example where they might be decoupled. I would prefer griddedCursorPoint then. >There are some actual examples of the this is a couple of commercial Mac >draw/paint applications, and MacApp explicitly supported this idea. When >gridding was on, the cursor jumped between grid points to be near the >mouse. This gave better feedback to the user about what would happen when >the button was pressed, since without it it's not as clear which grid point >would be chosen. The Morphic developers have chosen to do this at the Morphic side. I agree that griddedCursorPoint might be a worthwhile addition to InputSensor. OTOH, one could add a method to Point as in: sensor cursorPoint griddedTo: aPoint where aPoint is a normalized point which defines the grid in both x and y direction. All I'm saying is that there are only 11 senders of mousePoint (like Celeste). I would like those to use cursorPoint and then remove mousePoint. 226 clients of cursorPoint get 3 message sends to obtain the cursor point: cursorPoint sends mousePoint send primMousePt which is primitive. I want to keep primMousePt (it's categorized private anyway) and recode cursorPoint to use the primitive. That means that 226 clients suddenly have a very short message path for obtaining the cursor point: Cheers, Reinier.

1 0

Running Squeak on Linux
by Ivan Brusic 06 Apr '99

06 Apr '99

Hello fellow squeakers, I am unable to run Squeak on Linux. There are two different errors which I get when trying to run Squeak: 1) When using Ian's precompiled VM (SqueakVM-2.3-i386-linux-2.0.36) or using my own compiled VM (with the lastest sources) I get the following error: This visual class is not supported 2) When using Ian's other precompiled VM (SqueakVM-2.3-i386-linux-2.1.24) I get the following the following error: Could not open the Squeak image file 'squeak.image'. The file does exist and is correct. I have attempted to use relative and absolute paths (./squeak.image or /usr/local/squeak/squeak.image), but these attempts also do not work. Lex Spoon and I have noticed this behavior on Suns (running Solaris), but after several attempts, we manage to get those VMs to work. Has anyone come across this behavior? I would really to run Squeak, especially since I run only Linux nowadays (aka I corrupted my Windows partition :) ) My system is a PII 300 running SuSE 6.0 with the 2.2.4 kernel and glibc. Thanks in advance! Ivan -- Ivan Brusic | "Computers are nothing but a perfect Graduate Student | illusion of order" - Iggy Pop College of Computing & GVU Center | Georgia Institute of Technology | E-mail: ivan [at] cc.gatech.edu

2 1

upgrade766Sundry2-di problems
by Tim Rowledge 06 Apr '99

06 Apr '99

Loading 766Sundry2-di breaks the system - our old friend **System error handling failed** Tried loading into both a plain 2.3 (a _lot_ of upgrades to fetch!) and an image that had already got as far as 659. I don't have time to look into it any more today; dinner with friends to attend to. tim PS Interestingly apropos sigline that random number gen came up with.... -- Klingon Code Warrior:- 3) "Perhaps it IS a good day to Die! I say we ship it!" Tim Rowledge: rowledge(a)interval.com (w) +1 (650) 842-6110 (w) tim(a)sumeru.stanford.edu (h) <http://sumeru.stanford.edu/tim>

2 2

Printing
by Duane Maxwell 06 Apr '99

06 Apr '99

It seems to me that Smalltalk has always been extremely weak when it came to printing, and I was wondering if any thought had been given to improving this in Squeak. At the moment, I believe the entire printing infrastructure is based solely on Form>>primPrintHScale:vScale:landscape: <primitive 232> and nothing else. In these days of cheap, high resolution printers it seems rather odd that all we should get out is screendump quality stuff. Most of the VM implementations don't even do that - the Mac punts, and UNIX simply coughs out an XBM file. With the advent of Balloon and the primitive TrueType support, it seems at least possible to think about resolution-independent rendering, and defining some sort of abstraction for printing that could work across the majority of platforms. An approach that might be interesting is to play the same game that Apple's QuickDraw 3D for rendering. It recursively broke complex 3D objects into simpler datatypes, until the specified renderer would accept them. What might work for Squeak printing is something that queried the underlying OS printing architecture for its capabilites and decomposes the print job until it is accepted, or at worst, where it's at now - a Form. Another possibility is to include various translators for retained internal graphic objects to printing languages such as Postscript, but I think that it's not abstract enough. =================================================== Duane Maxwell dmaxwell (at) launchpados.com CTO http://www.launchpados.com Launchpad, Inc. (619) 578-8500 x226 Information contained herein is my personal opinion and not necessarily that of Launchpad, Inc. ===================================================

2 1

[REQ]Remove InputSensor>>mousePoint
by Reinier van Loon 06 Apr '99

06 Apr '99

Dan et al., while studying InputSensor I found that it has 3 methods to answer the coordinates of the cursor: a) mousePoint = ^ self primMousePt b) cursorPoint = ^ self cursorPoint c) primMousePt = <primitive: 90> mousePoint has 11 senders, cursorPoint has 226! and primMousePt just 1. Funny, the slowest method is used the most... I propose: a) to remove mousePoint; b) implement cursorPoint as <primitive: 90>. Cheers, Reinier.

2 1

RE: [REQ]Remove InputSensor>>mousePoint
by Norton, Chris 05 Apr '99

05 Apr '99

Hi Reinier & co., Although normally I consider method consolidation to be a laudable goal, I personally prefer readable selector names over selectors with abbreviations. In this case, I think that "mousePoint" is a lot more intuitive (from a reuse perspective (see Celeste, Scamper, etc.)) than anything with "prim" in it. Usually code that is labeled "prim" is reserved for the bowels of the system. Why not just copy the primitive code into the "mousePoint" method if you're concerned about speed? So, I guess I'm arguing to keep the "mousePoint" and "cursorPoint" selectors. Just tossing in my 2 bits... ---==> Chris PS> Thanks for helping to speed up Squeak! I applaud any and all efforts in that area!! :-) PPS> Does the single call that "mousePoint" makes to "primMousePt" really make much of a difference from a time to execute perspective? > -----Original Message----- > From: Reinier van Loon [SMTP:R.L.J.M.W.van.Loon@inter.nl.net] > Sent: Monday, April 05, 1999 3:58 PM > To: squeak(a)cs.uiuc.edu > Subject: [REQ]Remove InputSensor>>mousePoint > > Dan et al., > > while studying InputSensor I found that it has 3 methods to answer the > coordinates of the cursor: > > a) mousePoint = ^ self primMousePt > b) cursorPoint = ^ self cursorPoint > c) primMousePt = <primitive: 90> > > mousePoint has 11 senders, cursorPoint has 226! and primMousePt just 1. > > Funny, the slowest method is used the most... > > I propose: > > a) to remove mousePoint; > b) implement cursorPoint as <primitive: 90>. > > Cheers, Reinier.

1 0

Re[2]: Squeak and Sound Compression
by Jerome.Garcia＠wj.com 05 Apr '99

05 Apr '99

Asking the question from a position of almost total ignorance of both subjects, does someone know whether it would be possible to use genetic algorithms to help in the tuning of such parameters and filter coefficients? Jerome ______________________________ Reply Separator _________________________________ Subject: Re: Squeak and Sound Compression Author: johnm(a)wdi.disney.com at INTERNET Date: 4/3/99 3:28 AM >johnm(a)wdi.disney.com wrote: >With respect to John's call for PD codec sources, the problem is that so much of the good recent work has been at companies who want to license the technology as their business, so it's unlikely that they'd post a reference implementation to the web (or even publish a paper describing their techniques that >would make it easy for anyone else to make an external implementation). Sad but true. I do have a good book ("Introduction to Data Compression") that describes the general approaches used in MP3 and other modern codecs but I suspect that a great deal of careful work must go into tuning the precise parameters and filter coefficients to make a really great codec. -- John

2 1

Browser alternatives
by Bob Arning 05 Apr '99

05 Apr '99

A furher exploration of browser alternatives is available at: http://www.charm.net/~arning/BobsBrowser.5Apri1102am.cs.txt This one may hold greater appeal for category fans. This works in 2.3 (I am currently at change set 746). Cheers, Bob

1 0

[MVC]Tutorial on top views/controllers
by Reinier van Loon 05 Apr '99

05 Apr '99

People, I've put a small tutorial on MVC top views/controllers at http://www.2share.com/squeak/Tutorial-MVC-Top.st After studying this you should be able to: a) write your own custom top views; b) see the points of improvement in MVC ;-). Have fun, Reinier.

2 1

2.4 beta available
by Dan Ingalls 05 Apr '99

05 Apr '99

Folks - I have just put out image, changes and Mac VMs for 68K and PPC in a new folder named 2.4beta on the UIUC server. There is also a stuffit archive of all 4 there. Perhaps windows and unix supporters can make the analogous meterials available for their platforms. Disclaimer: I'm going out of town for three days, but I thought I would put this out so folks can use the time for testing. We have not had time to incorporate many of the submitted fixes and features. Please send email to me (not the list) if there are any you particularly want to see included. Thanks. Enjoy - Dan

2 2

Jump to page: