Squeak-dev March 1999

squeak-dev@lists.squeakfoundation.org

153 participants
338 discussions

WinSqueak issues
by mdr＠scn.org 04 Mar '99

04 Mar '99

Question: When you bring up the Squeak VM Preferences using the F2 key, it mentions Printer Setup. What kind of printing can I subsequently do, and how to do I do it? (Screen print?) Issue: Sometimes my Squeak screen goes white and it will not refresh. The cursor is still the Squeak standard one and moves, but I do not get any (now invisible) window/menu response from moving or clicking the mouse. My only solution is to click on the Win95 [X] window close button which quits Squeak. I can not reliably reproduce this, but it has happened several times (3?) when I have reexposed the Squeak window after it has been minimized for a while. I am using the standard 2.3 release VM. Anyone else seen this? Mike mdr(a)scn.org p.s. I also once got a stream of SetDBtoBits error message, but I think that was unrelated, and seemed to nicely go away after a bit of clicking around - scary but true.

1 0

Re: Linux 2.3 problems
by Marcus Denker 04 Mar '99

04 Mar '99

On Wed, Mar 03, 1999 at 03:50:21PM -0500, Richard L. Peskin wrote: > I have run into the following problems with the Linux (RH 5.2) version of > Squeak2.3. > > 1. The "Normal" cursor, the scroll bar between browser panes, the menu icon > (scroll region), adn other widgets are not clear. That is, details are not > showing up. For example if you move the cursor from the X window background > into Squeak, the "arrow" details get fuzzy. This occurs with X running on > the system console, but not with a remote X server. > Bert Freudenberg posted a fix to the list on Tue, 2 Feb 1999: |Subject: [BUG,FIX] CursorWithMask on X | |I really wanted to have that nice CursorWithMask but it showed only as a |black blob - stColorBlack and stColorWhite were not initialized with RGB |triplets (which are needed for 16/32 bpp visuals). A diff is attached. > 2. More important, the "default" object memory allocation is too small > (5m). While this allocation can be changed from the command line (-memory > option), how can the default be changed? > |Date: Mon, 1 Mar 1999 13:27:33 +0100 (MET) |From: Ian Piumarta <Ian.Piumarta(a)inria.fr> |To: markus_kohler(a)hp.com |Subject: Re: Installation help needed (Solaris, Linux) |Cc: squeak(a)cs.uiuc.edu | |> The default memory size that Squeak allocates is a bit low for the |> newest version ( we shoudl change that). | |sqXWindow.c(91): #define DefaultHeapSize (20*1024*1024) /* 20 megabytes */ | or: set the environment-variable SQUEAK_MEMORY, like (bash): export SQEUAK_MEMORY=20m Marcus -- Marcus Denker marcus(a)ira.uka.de fon@home:(0721)614235 @work:(0721)608-2749

2 1

RE: exception handling, differences from ANSI
by Raab, Andreas 04 Mar '99

04 Mar '99

> The ANSI standard exposes the situations in the programming > interface. It requires a separate class for each situation, which I think > is overkill. > This is not entirely true. The only thing that ANSI actually requires is that there are a couple of globals (e.g., MessageNotUnderstood, ZeroDivide) that respond to the right kind of messages. Whether these are instances of a single class or actual classes is irrelevant. Andreas -- PLEASE NOTE NEW EMAIL ADRESS! +===== Andreas Raab ========= (andreasr(a)wdi.disney.com) ==+ | Walt Disney Imagineering Phone: +1 818 544 5016 I I Glendale, CA Fax: +1 818 544 4544 I +======< http://isgwww.cs.uni-magdeburg.de/~raab >========+

4 4

Re: exception handling for Squeak available
by David T. Lewis 04 Mar '99

04 Mar '99

On Wed, Mar 03, 1999 at 12:35:13AM -0800, Craig Latta wrote: > > There have been 30 hits so far, but I haven't heard any feedback yet. > Well, it loads cleanly and the classes are well documented :) I'm not sure I can provide much useful feedback, other than to say thanks, and I hope that this helps make a T-Gen port possible.

1 0

Re: WeakKeyDictionary - does this code look right??
by Peter Smet 04 Mar '99

04 Mar '99

Thanks for your helpful replies. Andreas, Your example works fine, but if I change it slightly, the reference is no longer released. At what stage should the variable 'a' below get garbage collected? After I have run the code? After I have closed the Transcript/Workspace? Do I have to explicitly type a := nil? | w a | w _ WeakKeyDictionary new. a := Object new. w at: a put: 1. w at: Object new put: 2. Smalltalk garbageCollect. w inspect -----Original Message----- From: Raab, Andreas <Andreas.Raab(a)disney.com> To: 'squeak(a)cs.uiuc.edu' <squeak(a)cs.uiuc.edu> Date: Thursday, March 04, 1999 5:03 AM Subject: RE: WeakKeyDictionary - does this code look right?? >> Anyway, when I test the idea, the objects in the WeakKey >> Dictionary never seem to get garbage collected, even >> if there are no references to them (from the blocks, inspectors >> etc.) >> >What kind of keys do you use?! Also, having no references from an >*inspector* does not mean that there are no references to the object. If you >want to try if the stuff works at all then do something like: > > | w | > w _ WeakKeyDictionary new. > w at: Object new put: 1. > w at: Object new put: 2. > Smalltalk garbageCollect. > w inspect > >Given that you use the latest VM (e.g., 2.3 - the above code does *not* work >with pre-2.3 VMs) the Dictionary should contain 2 nil keys (e.g., the keys >have been collected). > >> What am I doing wrong? I am puzzled by the class comment in >> WeakKeyDictionary which states: >> >> "Clients are responsible to register my instances by WeakArray >> such that the appropriate actions can be taken upon loss of any keys." >> >See above - what you have after some keys are collected is *multiple* nil >keys in a dictionary. This can be extremely confusing and is potentially >dangerous. Therefore, WeakDictionaries should be registered by WeakArray >class so that they get informed that a key has possibly been removed. You >may want to look at WeakRegistry - it is doing this kind of action. > > Andreas >-- >PLEASE NOTE NEW EMAIL ADRESS! > >+===== Andreas Raab ========= (andreasr(a)wdi.disney.com) ==+ >| Walt Disney Imagineering Phone: +1 818 544 5016 I >I Glendale, CA Fax: +1 818 544 4544 I >+======< http://isgwww.cs.uni-magdeburg.de/~raab >========+ > >

1 0

ObjectMemory<SystemDictionary?
by Craig Latta 04 Mar '99

04 Mar '99

Hi-- What's the deal with the class variable SystemDictionary in ObjectMemory? There is no such variable now, but I assume there used to be one in the past, because of the line '"SmalltalkDictionary _ 8." "Do not delete!"' in ObjectMemory>>initializeSpecialObjectIndices. Is there any reason not to delete that line now, along with the corresponding slot in the special-objects array? thanks, -C -- Craig Latta composer and computer scientist craig.latta(a)netjam.org www.netjam.org latta(a)interval.com Smalltalkers do: [:it | All with: Class, (And love: it)]

1 0

GIF Animations
by Peter William Lount 03 Mar '99

03 Mar '99

Hi, I am wondering if anyone knows of any Smalltalk source for writing gif files. It would be incredible if the code also could create gif animations. All the best, Peter William Lount peter(a)smalltalk.org http://www.smalltalk.org/sjtrip.html

2 1

It's REALLY heating up out there...
by ohshima＠is.titech.ac.jp 03 Mar '99

03 Mar '99

Hello Squeakers. This mail may be off-topic here, but I can't help to imagine if Squeak runs on it:-) Sony announced the spec of the next generation PlayStation. According to the announce, it is something like: * 300Mhz new CPU called "Emotion Engine" (developped by Toshiba). - 128 bit registers, datapath, cache, etc. (the register can be used as two 64 bit registers) - two 64 bit integer unit, (I think 128 bit arithmetic is also possible) two floating vector unit, and MPEG2 decoder are integrated. - the integer arithmetic performace is 1/3 of 500Mhz Pentium III. - 6.2 GFLOPS floating arithmetic performance. - 32MB main memory. - the memory bandwidth is 3.2GB/s. (four times bigger than 64 bit * 100Mhz bus) * new graphics accelerator called "Graphics Synthesizer". - 75 million polygons/sec max. - 48MB/s memory bandwidth. * DVD-ROM. * USB. * PC-card slot. * IEEE-1394. It will be available this winter in Japan. People will buy it for a few handred dollars, and some may want to "create" animations, 3D models, and small programs on the machine. If there will be handy programming system running on it, it will be very useful. Squeak may be the candidate. The possible (half joking) future works for Squeak are: * 64 bit object format. * connecting Form/Canvas object with underlying hardware. * implementing VM using floating vector arithmetic:-) * provide simple interface for the language. * program editing by analog controller(s). OHSHIMA Yoshiki Dept. of Mathematical and Computing Sciences Tokyo Institute of Technology

9 10

Re: Fixes archive (was: entao)
by Bert Freudenberg 03 Mar '99

03 Mar '99

On Wed, 3 Mar 1999, John Brant wrote: > At 06:33 PM 3/3/99 +0100, you wrote: > >I have set up an automated "fixes collector" on the GSUG Swiki at > > http://swiki.gsug.org/SQFIXES > > > >Bugs: > >- Downloading of attachments except for plain text is not yet > >supported (you can, however, download the raw message to extract it by > >yourself). > >- The regexp I used for filtering is, umm, not really intelligent ;-) > > The mailing list archive http://squeak.cs.uiuc.edu/mail/squeak/ unpacks the > attachments. You might want to look at it. The messages are listed under > names such as http://squeak.cs.uiuc.edu/mail/squeak/msg06443.html where the > number is 2 less than the number in the mail's headers: "X-Mailing-List: > <squeak(a)cs.uiuc.edu> archive/latest/6445". Anyway, I thought you might be > able to hook this up to your script somehow... > Thanks, I didn't know that. Now, if I only could search your archive and the filenames of attachments would be preserved, we wouldn't need another list archive at all ... For now, I added a link to the message in the archive. /bert -- Bert Freudenberg Department of Simulation and mailto:bert@isg.cs.uni-magdeburg.de Computer Graphics http://isgwww.cs.uni-magdeburg.de/isg/bert.html Univ. of Magdeburg

1 0

Re[2]: exception handling for Squeak available
by Jerome.Garcia＠wj.com 03 Mar '99

03 Mar '99

I would like to second that "Thanks a bunch" and "Looks like a first class piece of work...". Just for fun, I have been trying to discover what would prevent porting of ControlWORKS and our semiconductor processing equipment software to Squeak from VisualWORKS (ignoring of course the sales job of convincing management to do such a thing). It looks like you have plugged one of the major holes. We use exceptions a lot. Now, if only there was a tool to help automate the port in order to give it a quick try :-) Jerome ______________________________ Reply Separator _________________________________ Subject: re: exception handling for Squeak available Author: "R. A. Harmon" <harmonra(a)webname.com> at INTERNET Date: 3/3/99 2:31 PM At 12:35 AM 3/3/99 -0800, you wrote: [snip] > I'll mark my exception handling framework (http://www.netjam.org/self/projects/smalltalk/exceptionHandling.html), as version one final on Monday, unless I hear about bugs before then. There have been 30 hits so far, but I haven't heard any feedback yet. [snip] Looks like a first class piece of work to me. I'm in the process of trying to relate your instance-based approach to ANSI's class-based. Very nicely done. Thanks a bunch. -- Richard A. Harmon "The only good zombie is a dead zombie" harmonra(a)webname.com E. G. McCarthy

1 0

Jump to page: