Squeak-dev March 2002

squeak-dev@lists.squeakfoundation.org

169 participants
610 discussions

Lego Mindstorm and Squeak
by Bergel Alexandre 04 Apr '02

04 Apr '02

Hello, I have finally make a little web page about my work with lego mindstorm. It works fine, but... I need help for getting out of my troubles with serial port. LegoMindstorm on Squeak here : http://minnow.cc.gatech.edu/squeak/2412 Do not hesitate to inform me about your comment... Best Regards, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._ Bergel Alexandre http://www.iam.unibe.ch/~bergel ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^

6 8

PPS: [FIX][OSProcess][Unix][VMMaker] OSProcessFix-sr v2.0
by Stephan Rudlof 03 Apr '02

03 Apr '02

Dear David, I have *very* good news! CommandShellV3-0 works now! If I type in an unknown command: $ ll ll: No such file or directory I have come to this result by incident: I was wondering why the *.image files are made executable and tried to execute one directly. And wonder: it works! It uses the same VM, but without some flags of my personal start script: sr@karl:~/Squeak/work$ cat `which mySqueak` #!/bin/bash SqueakProg=/usr/local/lib/squeak/3.2-4743/squeak SqueakDir=~/Squeak/work if [ $# = 0 ]; then SqueakImage=blocks.image else SqueakImage=$1 fi cd $SqueakDir exec $SqueakProg -memory 64m -xshm -nojit $SqueakImage OK, let's try the same script, but without '-xshm': Voila: this works! Reason for bug found! (Don't know what it is though...) Greetings, Stephan Stephan Rudlof wrote: <snipped> > But CommandShellV3-0 *crashes* for me for *unknown* Unix commands as stated > before for CommandShellV1-6: > > sr@karl:~/Squeak/work$ mySqueak standardVM > ioLoadModule(UnixOSProcessPlugin): UnixOSProcessPlugin: cannot open shared > objec > t file: No such file or directory > > Segmentation fault > > 1095321844 DisplayScreen>forceToScreen: > 1095321752 [] in DisplayScreen>forceDamageToScreen: > 1095321292 OrderedCollection>do: > 1095321168 DisplayScreen>forceDamageToScreen: > 1095321076 WorldState>forceDamageToScreen: > 1095173560 WorldState>displayWorld:submorphs: > 1095173468 PasteUpMorph>privateOuterDisplayWorld > 1095173376 PasteUpMorph>displayWorld > 1095172916 [] in WorldState>displayWorldSafely: > 1095173100 BlockContext>on:do: > 1095172824 BlockContext>ifError: > 1095172732 WorldState>displayWorldSafely: > 1095172640 PasteUpMorph>displayWorldSafely > 1095172548 Morph>refreshWorld > 1095160668 PluggableTextMorph>update: > 1095160576 ShellWindowMorph>update: > 1095160760 [] in Object>changed: > 1095160484 DependentsArray>do: > 1095160392 Object>changed: > 1095160300 CommandShell>endEntry > 1095160172 CommandShell>show: > 1095159896 [] in CommandShell>processError:count: > 1095159988 [] in Semaphore>critical: > 1095159804 BlockContext>ensure: > 1095159712 Semaphore>critical: > 1095159620 CommandShell>processError:count: > 1095158408 CommandShell>runProxy:signaling: > 1095158316 CommandShell>foreground:signaling: > 1095062568 CommandShell>doCommandsInString:ifTrue: > 1095062172 [] in CommandShell>doCommandsInString: > 1095062448 [] in BlockContext>newProcess > > Just to notify you: for me currently it is important that OSProcess > works..., and this works well! > > Thanks for your work, > > Greetings, > > Stephan > > Stephan Rudlof wrote: > > > > Needed for me to get UnixOSProcessPlugin compiled by postscript... > > Should work as it is and incrementally to OSProcessFix-sr v1.1. > > > > "Change Set: OSProcessFix-sr v2.0 > > Date: 12 March 2002 > > Author: Stephan Rudlof > > Preconditions: Squeak3.2gamma#4743, > > VMMaker-3-2-version5.5, OSProcessV3-0-dtl > > > > History: > > v2.0 upgraded for OSProcessV3-0-dtl; > > added: >>translateDoInlining:, removed: >>sandboxSecurity > > v1.1 UnixOSProcessPlugin>>sandboxSecurity always returns 0 for no security > > to be compatible with actual VMSources (don't know where exactly the > > incompatibility has arosen). > > v1.0 To work with VMMaker-3-2-version5.5, OSProcessV2-7 needs this fix. > > Otherwise the postscript doesn't work. > > " > > -- > > Stephan Rudlof (sr(a)evolgo.de) > > "Genius doesn't work on an assembly line basis. > > You can't simply say, 'Today I will be brilliant.'" > > -- Kirk, "The Ultimate Computer", stardate 4731.3 > > > > ---------------------------------------------------------------------------- > > Name: OSProcessFix-sr.3.cs.gz > > OSProcessFix-sr.3.cs.gz Type: application/x-gunzip > > Encoding: base64 > > -- > Stephan Rudlof (sr(a)evolgo.de) > "Genius doesn't work on an assembly line basis. > You can't simply say, 'Today I will be brilliant.'" > -- Kirk, "The Ultimate Computer", stardate 4731.3 -- Stephan Rudlof (sr(a)evolgo.de) "Genius doesn't work on an assembly line basis. You can't simply say, 'Today I will be brilliant.'" -- Kirk, "The Ultimate Computer", stardate 4731.3

11 25

[ENH][GOODIE] OSProcess V3.0 (OSProcess for Windows)
by David T. Lewis 03 Apr '02

03 Apr '02

[This is a repost - original did not make it to the list] OSProcessV3-0 David T. Lewis, 10 March 2002 Previous released version: OSProcess 2.7 Corresponding CommandShell version: 3.0 This is the first Windows release. Windows users please see 'Known bugs' below. Version numbers for OSProcess, CommandShell, and all plugins have been set to 3.0 for this release. See the separate CommandShell change set for a Unix shell window which uses OSProcess. OSProcess provides access to the external operating system from Squeak. Plugins are provided for Unix, Linux, and Windows systems. OSProcess can be loaded on other platforms as well, and placeholder classes are provided in the OSProcess hierarchy for other operating systems (Mac, OS/2, RiscOS), although support for these systems is not yet implemented. You can: - Open a inspector on the operating system process in which Squeak is running (OSProcess thisOSProcess). - Read or write on the stdin, stdout, and stderr streams (Unix only for this release). - Access the command line, and get or set environment variables (Windows: get environment only). - Fork and exec external programs, with control of the command line, environment, and stdin/stdout/stderr (Windows: run an external program). - Open an inspector on a child operating system process created from Squeak, with run state and exit status of the child updated asynchronously. - Run an external command: OSProcess command: 'xeyes' (Unix) OSProcess command: 'sol' (Windows) - Connect directly to input and outputs of external commands. See the separate CommandShell change set for a useful implementation, supporting commands like: (PipeableOSProcess command: 'ls -l *') output - Create pipes and connect them to external processes (see the CommandShell change set). - Fork an exact clone of the current running Squeak image (no image file) with parent and child Squeak processes proceeding from the same point in memory (UnixProcess forkSqueak). - Fork a Squeak using the disk based image (UnixProcess squeak). - Start a headless Squeak from the running Squeak image, giving it an expression to evaluate: (UnixProcess forkHeadlessSqueakAndDoThenQuit: [UnixProcess command: 'echo hello world']) - Tell the running image to go headless (UnixProcess decapitate), and reconnected the image with a window (UnixProcess recapitate). - Move the display to another computer on the network: (UnixProcess displayOnXServer: 'someOtherComputer') - Restart the VM on the fly (UnixProcess restartVirtualMachine). - Rebuild the VM, and restart Squeak with the new VM if the build was successful (UnixProcess makeVM). - Start your Swiki in a headless background process: (UnixProcess startSwiki: 'myswiki' onPort: 8081 loggingTo: 'log.txt') Known bugs in 3.0: - On Windows, OSPipe is implemented, but an OSPipe cannot yet be set for nonBlocking reads or writes. Reading from an OSPipe with no data available will block the Squeak VM (<ctl><alt><del> required!!). - On Windows, stdin/stdout/stderr access is not yet implemented. - On Windows, command pipelines for external commands are not yet implemented. CommandShell does work, and may be used for launching individual Windows programs as well as for the usual CommandShell internal commands and internal command pipelines. However, it is not yet possible to construct a pipeline of Windows programs. - Forking a child Squeak, as in UnixProcess>>forkSqueak, results in two instances of Squeak which share a single changes file. This has no practical impact if one or the other of the Squeak processes exits without doing very much, but it would probably result in a corrupted changes file if both Squeak instances do a lot of changes. I am leaving this as is for now, since a quick fix would require one of the Squeaks to save itself under an different image name. - The connectToXDisplay method is probably dangerous from a security perspective. - UnixOSProcessAccessor>>primitiveEnsetEnv uses the function unsetenv(), which is not available on Solaris (and possibly other systems). I do not know the portable equivalent of unsetenv(), so for now Solaris users will need to comment out the call to unsetenv() and rebuild the plugin. Other than failing the SUnit tests, commenting the function out has little practical impact. Changes in 3.0 since 2.7: - Added support for Windows. Windows programs can be launched. Squeak keeps track of its external programs and updates their runState and exitStatus. - Added a plugin for Windows. The plugin can run external Win32 programs, and provides notification to Squeak when an external process exits. It can create an OSPipe (but see bug notes above). Functions required for command pipelines and nonblocking pipes are not yet implemented. - Refactored the plugin classes to share code between Unix and Windows where possible. - Added WindowsThread to represent an OS thread of execution within an OS process. This may be generalized for other platforms as required (eg pthread support on Unix). - Did a general refactoring and renaming of classes to incorporate Windows support. Most importantly, class ConnectedUnixProcess has been replaced by class PipeableOSProcess in the separate CommandShell change set. PipeableOSProcess is platform independent, and collaborates with platform specific instances of ExternalOSProcess to eliminate the need for separate 'connectable' subclasses for each platform. The classes in OSProcess are: OSProcess ExternalOSProcess ExternalMacOSProcess ExternalOS2Process ExternalRiscOSProcess ExternalUnixOSProcess ExternalWindowsOSProcess ThisOSProcess MacProcess OS2Process RiscOSProcess UnixProcess WindowsProcess WindowsThread OSPipe (StandardFileStream) AttachableFileStream OSProcessAccessor MacOSProcessAccessor OS2OSProcessAccessor RiscOSProcessAccessor UnixOSProcessAccessor WindowsOSProcessAccessor (InterpreterPlugin) OSProcessPlugin UnixOSProcessPlugin UnixOSProcessPluginDynamicThisSession UnixOSProcessPluginInterpreterGetThisSession UnixOSProcessPluginNoThisSessionAvailable UnixOSProcessPluginStaticThisSession Win32OSProcessPlugin Win32OSProcessPluginDynamicThisSession Win32OSProcessPluginInterpreterGetThisSession Win32OSProcessPluginNoThisSessionAvailable Win32OSProcessPluginStaticThisSession

7 25

[ENH] vt100Telnet
by Felix Franz 03 Apr '02

03 Apr '02

A recent thread ("Squeak for everyday use??") reminded me of the telnet stuff I was working on. Well, here it is. The code is far from perfect, but it is enough to use vim :-) I filed it into a 3.2gamma image (#4743) and it worked for me. Cheers, felix from preamble: "Change Set: vt100Telnet Date: 27 March 2002 Author: Felix Franz improved TelnetClient. Use VT100TelnetClient open to open the Telnet window. Supports - full duplex mode (SUPP), (you can use screen oriented programs) - Sends the Terminal-type (TERM) to the host - Sends the Window size to the host (NAWS). The VT100 emulation is not perfect. It doesn't support scrollback."!

3 2

Dr. Dobb's
by David T. Lewis 02 Apr '02

02 Apr '02

Congratulations to Dan Ingalls and Adele Goldberg, who appear on the cover of the May 2002 "Dr. Dobb's Journal" as recipients of the 2002 Dr. Dobb's Excellence in Programming Award.

3 2

Annoyance: Hierarchical Browser
by Ross Boylan 02 Apr '02

02 Apr '02

Most of the time when I open a hierarchical class browser the focal class is invisible, since it is below the very tiny scroll area that is shown. This seems a bit odd, since that class is highlighted (when scrolled into view, and it is controlling what you see in the other windows anyway). I see this both on MS Windows and Linux, though I wonder if it's a font thing and it looks OK on, e.g., a Mac.

3 2

Why can't I do (aNonSmallInteger becomeFoward: aSmallInteger)?
by Brian Keefer 02 Apr '02

02 Apr '02

I'm guessing the reason is more involved than "because nobody asked", but such a capability would be nice for my dinky little LazyObject class. I suppose a reasonable workaround would be to force my SmallInteger into a Large(Positive|Negative)Integer, or make a wrapper for the SmallInteger. "All I need to know about Smalltalk I learned from Sesame Street" magicianMumford become: PeanutButterSandwich new.

9 13

Squeak shrink...
by Mayuresh Kathe 02 Apr '02

02 Apr '02

Hi all, I wanted to know the following: 1. How do I remove all traces of Morphic (I tried working using the code from MajorShrink, but it still leaves some traces) 2. How do I remove all traces of Balloon (same probs as above) 3. How do I change the menus so that they don't have items which try and access the old stuff 4. How do I get a clean system which doesn't have a lot of rif-raf? (in short, look like Twiggy, nut not feel the same ;) Also, I think, with my past comments regarding MVC, Morphic and Balloon, I have unknowingly offended some people. I would like to convey that I do respect their work (infact admire it) but what _I_ need is not what Morphic and Balloon can offer, and none of my comments were targeted to be offensive. Regards, ~Mayuresh

5 5

RE: Apple hyping java...
by David Chase 02 Apr '02

02 Apr '02

At 03:09 PM 3/30/2002 -0500, Bijan Parsia wrote: >Learning how to explore Squeak from the app into the code is a key >"aha" moment, in my experience. Obviously, you haven't gotten there >yet. It might be interesting to know what books/articles/whatever you used >so we can rectify them a bit. Didn't use any. Just picked it up, and started poking at it. Couldn't quite figure out where the source code was. I thought the flapping windows were a little weird -- if nothing else, it gives the idea that you can't really tell what you might or might not be able to do, until you actually poke it with the mouse. This meant, when looking for source code, that I was never really sure if I had looked properly (I am working from recollection here, it was some months ago). Eventually, I have better things to do. I will try again. (Much time passes.) Okay -- there is a bit of a theme to all the problems, and that is that the system is too tuned to expert use. Just for example, the tool tips take .8 second to appear. Why should a beginner be forced to wait to see these? So, I have a quest, which is to find that delay, and reduce it. There's insufficient cues. Over in the tools thing (that pulls out of the right border) there's no clear cue that what you are supposed to do is click AND DRAG. I ended up with some large number of funny yellow dots over there in the flaps, they must mean something, I can't make them do anything. And the world menu, why does it need a special "dismiss this menu" selection, instead of a little X in the upper left corner to make it go away? What's going on here? Why this inconsistency? There must be a reason, but it seems really ridiculous to have "dismiss this menu" spelled out as a choice. (And no, I don't think it would improve things much if you explained it to me. It would seem just as ridiculous to the next person who had not heard the explanation.) It's difficult to "find" things. I eventually decided that what I needed to do was find some message senders (for some message related to balloons, can't recall what). Is it "find"? Nope. Eventually, on a lark, I happened to right-click over in the unlabelled list that *happens to contain* what appear to be message names, and I discover the "senders of ..." choice. Now, I suppose, if I were playing Adventure, I would say "aha", but otherwise I'm just turning over rocks looking for pretty minimal treasure. So, after much looking, I find the method that contains the time for delay before display the help balloons, and change it to 100 ms. Now I'm ready to save my change. (Digression: A much less diplomatic rant was deleted, and I decided to substitute something more useful and perhaps productive. So, here I am, trying to replay my actions, and I CANNOT FIND the code that I found once already. That's not good.) So, I'm ready to save my change. That's save, spelled, apparently, a-c-c-e-p-t. At least, I hope that was right -- it asked me for my initials, and I wonder, in hindsight, if there was some sort of a try-it-out-before-committing button that I should have hit first (I mean, who knows, maybe I failed to turn over the UI rocks in the proper order?) This all seems a little terse, and it seems like a gratuitous change of jargon. Why not call all the finding things "find", and all the saving things "save"? Why not label the columns of stuff so I can know for sure what they contain? But now, with my faster help tips, I went poking through the music player, and selected something ("ratio") under the something menu on the waveform (light green) window, got a zero divide (why?) dismissed the exception notice, and now the waveform window is blank and useless. I wish I knew how to reset/restart it. Oh Well. Maybe the exception window went somewhere, and I should go looking for it. Nope, can't seem to find it. The learning curve here is really unpleasant. I can compare this with a couple of different things I've picked up in recent months. One was DrScheme, from Rice. I already knew Scheme, sure, but I didn't know the IDE (and I have the red, blue, and green Smalltalk 80 books to browse for Smalltalk examples, so I'm not completely without documentation). DrScheme had an example graphics thing I could cut from the doc, paste, play with. It was even a little bit buggy, so they could illustrate how to do something better (also in the doc). There's a "scripting" window in Squeak, but it doesn't appear to be what I wanted, since I cannot type in it, move it, close it, or change its size. Datapoint #2 is the Eclipse IDE for Java. It took me a while to figure out what was what, but I made some progress from the very start, and soon enough I knew enough to see that I would not have to take anyone else's word that it would eventually be worth my while to learn it. And yes, I am aware that Eclipse has Smalltalk roots, and I wonder if this turning-over-rocks sort of user interface is part of the Smalltalk style, because I have similar gripes about it. At one point, after using it for some days, it still took me *minutes* to find the proper button to push to get what I wanted. Please understand (and I can see how it would fit your model of the world not to believe this) I am actually trying to figure this out, and I am actually quite good at figuring things out. I've fixed Other People's Bugs in everything from Gosling Emacs, to device drivers in the LSI-11 FuzzBall OS, to a DVI-to-PDF converter in MikTeX, to David Gay's floating point-to-string conversion code. If it's not easy for me, you've got a problem. This feels like an Adventure Game, not a user interface. >I know, but I don't see why not. Squeak is great for writing a wide >variety of compliers. You've not done it, I take it. I've written a number of compilers, looked at many more. I've written a fair amount of code in Lisp and BCPL, and had quite good luck with some untyped C code that I wrote long ago (all roads lead to void *). I spent a good long time playing with functional languages (wrote an FP84 interpreter in BCPL, even). I'm not stupid, I'm not inexperienced. The compilers I was thinking of, and I should have made myself clear, seem to take at least 8 person-years to complete, and are commensurately large. It would be a tremendous win to write one of these in less time, but it would also be an expensive experiment. I don't know if they are large because they must be large, or if they are large because the use of a stupid language makes them large, or if static checking makes it possible for them to become large. Again, an expensive experiment to determine what's really going on, and I'm not really interested in ridiculous dogmatic answers from people who haven't even done it in the "wrong" language. You probably think I'm a neanderthal for using sucky Java, but most places I know of that are writing compilers, are using C or C++ -- they're not even using language with garbage collection. (Gcc, astoundingly, is written in K&R C.) They are not even willing to risk that minimal a change. What I don't see is what your purpose in arguing with me like this is. My suggestion/request, if you wish to do something about the adoption of Java where you think Squeak would do better, is to show off Squeak where it works well, and is clearly better than Java. If you wish to argue that Squeak is truly better even for writing compilers, you can say that, but as a compiler-and-other-things writer, I see Java's limits most clearly when working on things that are not compilers. I think you are making the job of promoting Squeak (to, for instance, Apple) unnecessarily hard if you insist on trying to show it off where Java does not most clearly suck. Swing's a pig -- it ought to be easy to make it look bad. Furthermore, nobody has ever said "wow!" about a compiler demo. David Chase

9 10

[BUG][VM][Unix][3.2gamma][SF][VMMaker] event handling broken
by Stephan Rudlof 01 Apr '02

01 Apr '02

Generating a recent Linux 3.2gamma VM with VMMaker-3-2-version5.5.cs ConsolidateTo4744-JMM.4.cs 3.2-4743 #1 [oss audio, xshm] Tue Mar 5 17:50:35 CET 2002 gcc 2.95.4 actual SF sources gives a bad event handling: mouse button presses only work after *moving* the mouse thereafter! If you click e.g. a collapse button it becomes darker, but the window collapses just after moving the mouse then. This is new compared with older VMs. Greetings, Stephan -- Stephan Rudlof (sr(a)evolgo.de) "Genius doesn't work on an assembly line basis. You can't simply say, 'Today I will be brilliant.'" -- Kirk, "The Ultimate Computer", stardate 4731.3

5 9

Jump to page: