Squeak-dev January 2002

squeak-dev@lists.squeakfoundation.org

176 participants
612 discussions

demos and Re: [M206] how to do M206 diagrams in Squeak
by Ross Boylan 03 Jan '02

03 Jan '02

Ah-- I loaded ConnectorMorphStates-nk.002.pr, not the NK- project you mentioned below. See below for some comments on demonstrations. At 11:04 PM 1/1/02 -0800, Ned Konz wrote: >On Sunday 30 December 2001 07:12 pm, Ross Boylan wrote: > > >I can read in the connectors project from BSS, but when I try to load > > >http://nedkonz.dhs.org:8080/Ned/uploads/12/M206.pr > > > > I get a message asking me how to convert ncButtonBar (from memory) to a > > "modern class". > >Note that the M206 project doesn't do much that the BSS connectors project >doesn't already do. Just a couple of customized shapes you probably don't >need. Note that the button bar examples that come with the Connectors Demo >project are pretty much official UML; the M206 stuff isn't. I wouldn't mess >with it except as an example of how to make your own shapes. > >Hmm... which BSS project did you load (I think you need the >NK-ConnectorsDemo.056.pr version)? It should have the NCButtonBar... >It loads ok for me... > > > Here's the method it asks me to fill in > > nCButtonBarbosfcebb0 > > > > ^ PutNewClassHere > > > > I'm at 3.2alpha #4599 > > > > By the way, what is M206? Some school class, or is it something more > > august? > >Someone was taking an online class and wanted to do diagrams using some funky >diagramming scheme. I posted an example to the list. > > > P.S. Thanks for the tip on EventRecorder; it saved putting a lot of work > > into something that wouldn't work. > >Great. Did you find something that would work? Plain old text. By the way, should I be able to bold face in a workspace? I can change the point size only with the menus I can see. I was thinking of using some diagramming tool to describe the project internals, but that would force me to add that project to mine, so I left it out. .. Hmm, that makes me think: perhaps I could make the diagrams a separate project, and say, first you must file in your project to get it to work. It will be nice when the new modules stuff is available. It would be nice if squeak had a little demo system built in. If you could take the event recorder scripts and then convert them into messages to windows (i.e., instead of capturing the click you capture the message activating the menu item) it would be nice. Probably a highly non-trivial exercise. I know I worked with automated testing tools for C++ graphic applications that did something like this extremely poorly. With smalltalk it should be much easier, but that doesn't make it easy. Or perhaps the application could be script-enabled in some fashion. I haven't looked at the scripting stuff enough to know if that's plausible.

1 0

[ANNC] ComSwiki 1.2 ("Harvest")
by Jochen F. Rick 03 Jan '02

03 Jan '02

Greetings, I am hereby announcing Swiki 1.2 ("Harvest"), running under Comanche 4.11. You can get pre-packaged downloads from the Swiki Swiki: http://minnow.cc.gatech.edu/swiki/15 It contains some major changes, including support for Squeak project pages and local administrator form control. For a limited time, you can play with the new features on manatee: Improvements: http://manatee.cc.gatech.edu:8080/refs/9 Projects/forms: http://manatee.cc.gatech.edu:8080/refs/13 http://manatee.cc.gatech.edu:8080/refs/14 Because it is a major revision, there is a good chance of bugs that will need to be corrected in Swiki 1.3. Please send all bug reports, comments, and random gunfire to swiki-bugs(a)cc.gatech.edu and/or pws(a)cc.gatech.edu Thanks for your consideration, Je77

1 0

making the mac startup faster...
by John M McIntosh 03 Jan '02

03 Jan '02

I've been looking into this. I've noticed on windows why when you startup Squeak it just snaps up! But it takes seconds on the mac, why? Thanks to some OS-X tools, let alone the usage of fs_usage I've noted that 3.2.1B5 does 1,574 I/O operations in order to startup. By apply some thought to how plugins are looked for I've cut that to 902 I/Os Right now I get timings of 11ms to start main to the point of image reading 1,852ms to read a 15,1903,04 file, 7.8MB sec? 386ms to fixup image. 1,250ms to first update screen event in the interpreter. I'm considering mmap and other interesting things to perhaps speedup the read, along with more review of how the directory lookup logic and directory entries logic works. (Somewhat ugly due to the transition to os-x). Let alone how the startup code hunts for the changes and source file, and why we need to tap the nonexistent applescript plugin at startup time. So I'm wondering if there are some suggestions out there. -- -- =========================================================================== John M. McIntosh <johnmci(a)smalltalkconsulting.com> 1-800-477-2659 Corporate Smalltalk Consulting Ltd. http://www.smalltalkconsulting.com ===========================================================================

4 4

Re: [ANN][VM][SF] my Unix patches integrated into SF layout
by Lex Spoon 02 Jan '02

02 Jan '02

> > But it doesn't work for me (yet). Drats. Okay, let's see.... > - First, I expected to have to run the copied version of configure in > the build directory as with the current SF tree; but ok, you documented > that one should use the one in "platforms/unix/misc" > - Then, I see that all the *.c files are compiled from the "platforms" > hierarchy, and not in the copied hierarchy under the build dir; here it > really doesn't make any sense to copy everything first It's just as easy to use the files right from platforms/unix/misc, as to remember to copy them over. Also, while *developing* these files, it is very nice to be able to edit them and the re-run the compile, without having to also copy them again. So, I left them there.... > CASE(0:) Drats -- different awk's seem to do handle substr() differently -- just try the output of this under gawk versus mawk: gawk 'BEGIN {print substr("12:",0,2); }' Here's a gnuify that uses a simpler and hopefully more portable trick (why didn't this come to mind before?!): do (($NF) + 0) to read off the initial integer of a sting. Ugly? Yes, but it's very near to the heart of awkdom -- if it doesn't work, I'm not sure it can really be called "awk". "ack", "ptoey", "gross", "spglark".... but not awk. Could you give it another try? Below is a new file, and I've updated sha.tgz on the Swiki. -Lex ==== #!/bin/awk -f # # Copyright (C) 1996 1997 1998 1999 2000 2001 Ian Piumarta and individual # authors/contributors listed elsewhere in this file. # All rights reserved. # # This file is part of Unix Squeak. # # This file is distributed in the hope that it will be useful, but WITHOUT # ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or # FITNESS FOR A PARTICULAR PURPOSE. # # You may use and/or distribute this file ONLY as part of Squeak, under # the terms of the Squeak License as described in `LICENSE' in the base of # this distribution, subject to the following restrictions: # # 1. The origin of this software must not be misrepresented; you must not # claim that you wrote the original software. If you use this software # in a product, an acknowledgment to the original author(s) (and any # other contributors mentioned herein) in the product documentation # would be appreciated but is not required. # # 2. This notice may not be removed or altered in any source distribution. # # Using or modifying this file for use in any context other than Squeak # changes these copyright conditions. Read the file `COPYING' in the base # of the distribution before proceeding with any such use. # # You are STRONGLY DISCOURAGED from distributing a modified version of # this file under its original name without permission. If you must # change it, rename it first. # Author: Ian.Piumarta(a)INRIA.Fr # # Last edited: 2000-10-18 20:30:52 by piumarta on emilia.rd.wdi.disney.com # Rewrite the interpreter source in $1 to use GNU C extensions, writing the # modified file to stdout. Author: Ian.Piumarta(a)inria.fr # 2001-8-1 Updated by Lex Spoon to (hopefully) work on all awk's # 2002-1-1 Updated by Lex Spoon again.... BEGIN { print "/* This file has been post-processed for GNU C */\n\n\n"; # print "copying first section of file" > "/dev/stderr"; stage= 0; } /#include "sq.h"/ { print "#include \"sqGnu.h\"\n"; next; } # An inlining comment means the interpreter was not inlined /inline: true/ { # print NR, $0 > "/dev/stderr"; # print "" > "/dev/stderr"; # print "*** interpreter was not inlined: bailing out! ***" > "/dev/stderr"; # print "" > "/dev/stderr"; stage= 1000; } (stage == 0) && /^int interpret$void$ \{/ { # print "interpret: adding static register assignments" > "/dev/stderr"; stage= 1; print; next; } (stage == 1) && /^ char\* localIP;/ { print " register char* localIP IP_REG;"; next; } (stage == 1) && /^ char\* localSP;/ { print " register char* localSP SP_REG;"; next; } (stage == 1) && /^ int currentBytecode;/ { print " register int currentBytecode CB_REG;"; next; } (stage == 1) && /^$/ { # print "interpret: adding bytecode dispatch table" > "/dev/stderr"; print " JUMP_TABLE;\n"; # print "interpret: rewriting case labels and outer breaks" > "/dev/stderr"; stage= 2; FS=" "; next; } (stage == 2) && /^ case / { print " CASE(" (($NF) + 0) ")"; next; } (stage == 2) && /^ break;/ { print " BREAK;"; next; } (stage == 2) && /^\}/ { stage= 3; print; next; } (stage == 3) && /^int primitiveResponse$/ { print; # print "primitiveResponse: adding primitive dispatch table" > "/dev/stderr"; print " PRIM_TABLE;\n"; # print "primitiveResponse: rewriting case labels" > "/dev/stderr"; stage= 4; FS=" "; next; } (stage == 4) && /^ switch \(primitiveIndex$ {/ { # print "primitiveResponse: adding primitive dispatch" > "/dev/stderr"; print " PRIM_DISPATCH;"; print; next; } (stage == 4) && /^ case / { print " CASE(" (($NF) + 0) ")"; next; } (stage == 4) && /^\}/ { # print "copying last section of file" > "/dev/stderr"; stage= 5; FS=" "; print; next; } # default { print; next; }

2 1

Re: VMMaker thoughts
by Lex Spoon 02 Jan '02

02 Jan '02

> > First, it doesn't really need to copy files around, does it? It's no > > trouble to write the makefile, at least on Unix, to compile files > > directly from the two different locations. > This may well be true on unix, where the autoconf stuff allows the > makefile to be generated depending on the configuration. The hacking > about that I did basically made it look for any subdirectory of > 'plugins' and compile each one as an external plugin. I imagine one > could make it find each 'plugin' sub-dir and then look back across to > the platforms/unix/plugins/{matching plugin name} in order to save > copying the files. Yes, this is what I meant. I've uploaded a new copy of SHA now that works like this, i.e., files are pulled directly from platforms/unix and platforms/Cross. Also, if you have files "intplugins" and "extplugins", the script will use those to decide which plugins to compile internally and externally. I haven't heavily tested the internal case, but it looks at least close. By the way, SurfacePlugin has SurfacePlugin.c in the Cross directory. This fouled up my script, which isn't smart enough to deal with files being overridden.... > Other platforms do seem > to need the files copying, and thus making a special case for the unix > seemed pointless at the time. What I'd _really_ have liked is to use > links, but not many platforms do links in any sensible manner and there > is no Squeak interface to support them anyway. I don't understand why any platform would care whether the subdirectories were all merged together, but I'll admit, it's not a terribly big deal. The Unix port's makefile can just ignore the copied files and use the originals directly, and the other ports can use the copied files, and everyone can be happy. > > Finally, let's have some space in CVS for files that are shared among > > multiple platforms. > It's already there; it's called the 'Cross' platform in the tree. sq.h, > sqNamedPrims.c etc live there as do assorted plugin files like most of > the plugin.h files. Note that files in the Cross platform can be > overridden by the ones in the actual platform. As an example, there is a > default copy of sqConfig.h that would just make your compiler complain > if it were not overwritten. One thing I must try to get around to is > making an example platform loosely based on the old sqMacMinimal.c file. Ah, indeed libmpeg is in there. Silly me. > The nice part about copying files is that it results in a complete > source tree that you can pickle and save for archival purposes or send > to somebody else. Good point, that's kinda nice.... > The baddest part is that it is tempting to edit the > files in src/ and recompile rather than editingthe copy in /platforms > and rerunning VMMaker. I think there must be some decent heuristic to > check the datestamps and warn about overwritten files, maybe even suck > them back - but I haven't sat down and worked it through yet. Part of > the problem is that old bugbear of poor file access methods. You know, make is pretty good about this kind of thing. Let it know the original location, and the copied location, and if the timestamps of the original is newer, then do a fresh copy. Then again, if ou could figure a way to tell make where the original file was, then you wouldn't need the copy, anyway. :) > > > In particular, libmpeg should be shared; it seems > > silly to copy all of libmpeg to each platform's MPEGPlugin subdirectory. > It is, and it isn't .... I don't understand. The files in platforms/Cross/Mpeg3Plugin *don't* get copied? Or is it only subdirectories that don't get copied? Or what? My picture was that src/Mpeg3Plugin would be the union of: 1. Mpeg3Plugin.[ch], generated from the image 2. platforms/Cross/Mpeg3Plugin/* 2. platforms/foo/Mpeg3Plugin/* What am I missing? > ... and it needs to be added to the SF repository and the unix makefile > stuff seems to need some update to make it make it. Lex, didn't you do a > hack to makefile.in that made it work for unix? Care to merge it into > the SF version of the make files? I'll do it whenever I find a few spare minutes; it should be pretty close already.... I wouldn't mind being beaten to it, either, by the way. It'll take updating the mkMakefile scripts and re-running them. > PS the explanation of the Cross platform etc is in the help window of > the VMMakerTool. Or at least, it should be. I was reading the swiki page, and running VMMaker by hand. I'll have to grab VMMakerTool, but the swiki it is on seems to be unavailable right now. -Lex

3 2

[GOODIE] Multiple line printing in inspectors
by Leandro Caniglia 02 Jan '02

02 Jan '02

>From preamble: Change Set: fullPrintOnInspector Date: 28 December 2001 Author: Leandro Caniglia Allows objects to be printed using several lines when seen from inspectors. Rationale. While it isn't a good idea to embed CR characters in printOn: methods, some objects require multiple lines to show a meaningful description of themselves. Since inspectors provide a whole panel to display such descriptions, there is no reason to use a one-line string with them. This change set replaces the use of printOn: in inspectors with the fullPrintOn: message. An example is given in PopUpMenu.fullPrintOn:. /Leandro

1 0

Re: [ANN][VM][SF] my Unix patches integrated into SF layout
by Joern Eyrich 02 Jan '02

02 Jan '02

I like the "+ 0" awk trick... Works like a charm now. > > - First, I expected to have to run the copied version of configure in > > the build directory as with the current SF tree; but ok, you documented > > that one should use the one in "platforms/unix/misc" > > - Then, I see that all the *.c files are compiled from the "platforms" > > hierarchy, and not in the copied hierarchy under the build dir; here it > > really doesn't make any sense to copy everything first > > It's just as easy to use the files right from platforms/unix/misc, as to > remember to copy them over. Also, while *developing* these files, it is > very nice to be able to edit them and the re-run the compile, without > having to also copy them again. So, I left them there.... Yes, that's right. I'm just bad at remembering exceptions, so I just had to complain after I had barely learned the VMMaker rules... ;-) I noticed that sq.h is taken from the build dir hierarchy, by the way, so it's not entirely consistent... Or at least I don't understand what is pulled from where, exactly. And I get warnings about duplicate rules in the Makefile. I'm not courageous enough to dive into this autoconf/Makefile stuff to find out why or do something against it, though... However, I have attached a new version of sqUnixMemory.c, which removes a few pointer/int cast compiler warnings. Briefly tested with/without mmap enabled. Bye, Joern

1 0

Re: finalization and garbage collection
by Lex Spoon 02 Jan '02

02 Jan '02

John M McIntosh <johnmci(a)smalltalkconsulting.com> wrote: > >Is there a method that is guaranteed to be executed just before an > >object is garbage collected? > >I thought #finalize was it, but looking at senders in the image and > >ObjectMemory code, it appears that this only gets called when > >something that is weakly held gets cleaned up. > > > >The weak apparatus seems to use both finalize and the executor > >mechanism, and I'm not sure what the roles of these two parts are. > > Here's my understanding. First, only objects that are registered in certain collections will get finalized. It's not automatic. Look up the way, say, sockets arrange to get finalized, to see how this works. Second, finalization happens in Squeak *after* the object has been collected. Thus, barring fancy magic stuff, a different object must be used to do the finalization. That's the "executor", and by default it's just a shallow copy. A shallow copy will look just like the original object, so it's easy to forget that it's not precisely the same object. This would make a great Swiki page, for someone who really understands it... -Lex

1 0

RE: Squeak Tutorial for Java programmers
by Ivan Tomek 02 Jan '02

02 Jan '02

Thanks Gio, A minor complaint - is there any reason you spell Smalltalk as SmallTalk? Ivan > > > I have posted a new version of the tutorial on my site: > <http://objectsroot.com/squeak/squeak_tutorial.html> > > Please give a look and send me bugs reports. > I will expand soon some sections. > ciao ciao > Gio > >

1 0

Squeak Tutorial for Java programmers
by Giovanni Giorgi 02 Jan '02

02 Jan '02

I have posted a new version of the tutorial on my site: <http://objectsroot.com/squeak/squeak_tutorial.html> Please give a look and send me bugs reports. I will expand soon some sections. ciao ciao Gio

1 0

Jump to page: