Squeak-dev January 2000

squeak-dev@lists.squeakfoundation.org

131 participants
468 discussions

RE: Type Safety (was Re: fun and empowerment)
by Lex Spoon 29 Jan '00

29 Jan '00

> > Largely a myth. However, I am convinced that "design by contract" and > similar methods can be worth the time and effort and could easily be > supported by Smalltalk. Has anyone written an Eiffel-like assertion package > for Smalltalk classes? SUnit has largely the same feel: you write checks in parallel with the code. The checks in SUnit are external to the classes, though. Indeed, I've found that not only do I catch a few bugs by using SUnit, but that I end up with cleaner code that I understand better. I think it's just from having to sit and read through the code carefully which does this--making up a good check means you have to really think about how the code works. For internal checks like you describe, it would be cool if Squeak had a global assert: function: Object assert: aBlock "check that aBlock is true. Can be disabled via a preference." Preferences doAssertions ifTrue: [ aBlock value ifFalse: [ self error: 'Assertion failed' ] ] Lex

1 0

[ENH] Squeak functionality speed and safety improvements
by Michael Rutenberg 29 Jan '00

29 Jan '00

Attached is a change set with a number of improvements to Celeste. I think they make it easier to use, and more robust to failures. I have incorporated lex's feedback and erred on the side of safety. I've tested this and have been using it with my personal email for the last week or so. Enjoy! Mike mdr(a)scn.org ____ Change Set: CelesteImprov-mdr Date: 28 January 2000 Author: Mike Rutenberg (mdr(a)scn.org) A number of fixes and improvements to Celeste. (1) Ensure the definitive email database is on disk over a snapshot, so that database changes are reflected next time the image is started (a) Write database to disk before snapshot (b) Read database from disk when Celeste next receives modelWakeUp (2) Add Balloon Help for buttons and Celeste specific menus (3) Check database integrity when opening by ensuring the modification times of the files corresponds to the saveDB save order (4) Remove a redundant (and costly) self changed: message from updateTOC. When displaying a new/modified large category (e.g. the 8000 Squeak messages of last year) it cuts the update time almost in half from 14 seconds. A variety of smaller fixes including: (5) MailAddressTokenizer would go into an infinite loop when evaluating email addresses containing accidental control character white space. Change the value of class variable CSNonSeparators to handle this. (e.g. MailAddressParser addressesIn: (22 asCharacter asString, 'testuser(a)host.com')) (6) Fix From filter and Subject filter buttons to correctly display their state when you decide to cancel entering a new filter (7) Saving mail now gives visual feedback (8) Move Again button fixed to operate gracefully when no message is selected (9) Minor support changes to rest of system Defines View|setBalloonText: aString for morphic compatibility for Buttons Cosmetic chage to an error message in FileDirectory Change to ensure modelWakeUp is sent to the active window after Snapshot-and-continue (10) Automatic saves of the database following network communication is now controlled by an flag and have been moved slightly to ensure that slow writing will not time-out network connections. Default behaviour is to do save, as previously.

1 0

Re: Type Safety (was Re: fun and empowerment)
by garau＠ar.ibm.com 29 Jan '00

29 Jan '00

> reason I mention it here is the idea that you can run Dylan in two modes: > a dynamic, interactive, prototyping mode for development, and an > efficiently compiled, static mode for delivery. Now for the $400,000 > question: how difficult would it be to add that sort of capability to > Squeak? :) We have already started ! With Concrete Type Inference you implicitly build a call graph. This is very easy to do in a procedural language where the name of each procedure leads to the real implementation. But it is more difficult in a language with dynamic dispatch (also known as late-binding). In Squeak a message send doesn't tell you which method will be invoked. You need to know the type of the receiver first. If after analyzing the whole program (a starting expression) you got only one particular receiver type, then you know the method that will be invoked at run-time. You can inline the method or replace the late binding with a static binding. Cool!! This is even better that what is done with C++. If you define a function as virtual, you will always get late-binding through the virtual table. Having the call graph, allows you to take only the classes and methods that would be called in *any* execution of that program. This is called 'application extraction' and with this in hand a *real small* 'hello world' application would be possible. So, I am confident that an efficent and compact delivery mode is feasible in Squeak. The firsts steps are taken (take a look at what is done in http://typeinference.swiki.net). However, this is not an easy task, and there are some very interesting problems to be solved. Anyone interested in this subject should read: - Ole Agesen Phd Thesis. (downloadable from http://typeinference.swiki.net) - David Grove Phd Thesis. ( http://www.cs.washington.edu/research/projects/cecil/www/Papers/papers.html) Francisco (Pancho) Garau Ph: (5411) 4319-6213. Fax: (5411) 4319-6422 garau(a)ar.ibm.com (work) - - - fgarau(a)bigfoot.com (home)

1 0

[ENH][Experimental] Password free access to your email
by Michael Rutenberg 29 Jan '00

29 Jan '00

I would suggest not including this in the main distribution at this point. If anyone has any problems using this, please let me know and ideally tell me the name of your POP3 server. And if it is useful, also let me know. I could always get it to work around disfunctional POP3 servers. Mike mdr(a)scn.org ___________ Change Set: POP3authentication Date: 29 January 2000 Author: Mike Rutenberg (mdr(a)scn.org) Implementation of the POP3 APOP command to allow authentication to your mail server without sending the password over the network (RFC 1939). Requires the MD5 class earlier submitted by Duane Maxwell (Thanks! You saved me lots of time). NOTE: These changes work fine, but should be considered experimental for now. On one server I tried, it appears that the POP3 server does not respond gracefully and correctly to the APOP command, and rudely closes the connection thus preventing you from normal authentication.

1 0

Re: Playstation project
by Tansel Ersavas 29 Jan '00

29 Jan '00

I am happy to see interest in Squeak-AIBO port. This summer I have acquired a few AIBOs (ERS 110) for the purpose of preparing for the legged robots competition where AIBO is the official platform. Although AIBO is promoted as an open platform it is not the case at all. There is development software (SDK and compiler) for the AIBO available to qualified ROBOCUP competitors but: a) It is $3500 apiece and you have to get one for each CPU (min 4) b) You have to sign a tough non-disclosure agreement c) You cannot use the development platform for any other purpose other than the competition. Obviously releasing a Squeak for AIBO under these terms is totally out of question. Even porting to a platform like that would not be allowed under Squeak license, as I understand it because you can't release it to the public under the terms of the NDA. After all this I was disillusioned and put the project on halt. BTW again my interpretation of the Squeak license would not allow for porting to proprietary hardware that a company wants to keep the details a trade secret and therefore would not want to disclose the ported code as it would give away a lot about the hardware. Is this a correct interpretation? There is not much doc about Aperios (some on its predecessors) but it is an interesting OS esp. for such purposes. It would not be trivial to port Squeak on it, but would be well worth it. I have a cool system that I have partly ported to Squeak. It is a merger of fuzzy logic and object oriented approaches that gives you a lot of power in controlling things like AIBO and it would do a better job of representing emotions and certain movements than what they have currently. It is not ready to be released yet and without such platforms it is difficult to demonstrate the power of that system but Luciano Notarfrancesco had a quick glimpse of it and you can ask him what he is thinking about it. Hardware wise AIBO is more than enough to run a headless Squeak with my extensions quite comfortably. We were able to cram a lot to a Cassiopeia E100 (16M CE handheld) for a demo that had about 100+ full color images occupying more than 8+M of memory and 2MBs of sound (For the most part, decompression on the fly did not prove practical for CE for us). In AIBO, the most space that we may want to occupy will not be the code logic, but things like sounds. Anyway, if Sony continues with this policy I might need to change the computer and software in AIBO and put my own embedded system with Squeak in it. I hear Alan has some sort of contact with Sony and he could be in an influencing position. Is this the case Alan? Maybe you could help us with PS2 and AIBO development platforms. To friends in Japan: if you are working in Sony or have friends in Sony may have a better chance of raising your voice to keep to their word and release information and development tools about OPEN-R and Aperios, or just remove those OPEN-R stickers from them. Also, I will be in Tokyo on 2nd or 3rd of February briefly, if there is a chance of meeting you can email me privately. Cheers Tansel Yoshiki Ohshima wrote: > > Hi Soren and Luciano, > > # Thank you for sending the mystery fish movie the last > # time! > > Soren wrote: > > It seems that Sony is supporting at least one open source > > project with hardware and development kits. See > > http://wulfstation.sourceforge.net/ . The goal is to make a > > Beowulf supercomputer out of a cluster of networked > > Playstations. So maybe we will not have to pay. Certainly > > our project is as worthy as theirs. > > Interesting. > > One thing I can say is wulfAIBO is less worthy than > wulfstation. > > # theirs? > > > **(Surely the hardware accelaration part will take much much longer:-) > > > > That is also what Luciano thinks, but I disagree. Imagine > > porting Balloon3D to a machine which required OpenGL > > commands to be issued to the hardware. It would not be > > difficult, would it? Surely the PSX2 comes with a C library > > for accessing the rendering hardware -- not OpenGL, of > > course. > > What I meant was it will take longer to write the 3D > support code than to write minimal interface support such as > BitBlt and pointing device. I don't say 3D part is easy, > but I agree with you that what to do is not uncertain. > > # But my experience taught me that making it work and tuning the > # performance is quite different. > > ## I guess we might want to implement it so that rendering > ## and interpreting are carried out in parallel. > > Luciano wrote: > > AIBO! Having Squeak running on a dog! Great idea!!! I > > love it!! > > Do you have an AIBO? A friend of mine (Squeaker too) > > has one, and I might be able to play a little with it. > > But I don't have a development kit for AIBO, actually > > I'm not sure such a thing exist. Both PSX2 and AIBO > > have MIPS in them, right? > > I don't have one, but a research group I sometimes visit > own one. AIBO has (built-in) "motion editor" and you can > program its motion for some extent, but I don't think there > is a development kit available. > > They say the chip on AIBO is R4200. The OS of AIBO is an > "object-oriented OS" Aperios, which was formerly called > Apertos, which was formerly called Muse. The OS was designed > by Yokote-san based on his Concurrent Smalltalk work. > > The chip on PlayStation2 is *completely* new. I'm not > sure that even it has 32 bit arithmetic and 32 bit > load/store ops. > > -- Yoshiki

1 0

Re: xml.smalltalk.org and mod.smalltalk.org
by Peter William Lount 29 Jan '00

29 Jan '00

Hi David, There is a Smalltalk XML parser, although some ports are buggy, from Indelv.com (link via http://xml.smalltalk.org) that you could play with. There is no squeak version yet? After looking at it please let me know what do you think of it? All the best, Peter William Lount peter(a)smalltalk.org http://www.smalltalk.org ---------- From: David Cramer <dacramer(a)videon.wave.ca> To: squeak(a)cs.uiuc.edu Subject: Re: xml.smalltalk.org and mod.smalltalk.org Date: January 28, 2000 5:38 PM I'd be very interested in playing with a Squeak-based XML parser. I have been concerned for quite some time with the seeming lack of XML-related Smalltalk activity. Back in October, after just attended the XML World Conference in Ottawa, I posted a message to this mailing list regarding a conversation I had had with Dave Thomas, head of Object Technology International (OTI, now an affiliate of IBM Canada), and with over 30 years of experience in the computer industry, elected in 1997 to the IBM Academy in recognition of his pioneering work on object technology. Thomas confirmed my suspicions when he said matter-of-factly that Smalltalk was inherently superior to Java for the kind of programming needed for large-scale XML data management, but that Java had industry buy-in regardless of its immaturity and shortcomings. Unfortunately, he suggested it might take a couple of years for the realization to hit. I have a strong instinctive feeling that the Squeak community can play a pivotal role in the confluence of markup language technologies and Smalltalk programming, to everyone's benefit. I've been learning Smalltalk without a specific goal. An XML-related project in Smalltalk for me would be something I could really devote some time to. Good luck with InDelv! Regards, David At 2:09 PM -0800 1/28/00, Duane Maxwell wrote: > >Incidentally, we've ported the "open source" InDelv XML parser ><http://www.indelv.com> , but we're not quite sure what to do with it. >It's based a lot on the Java/DOM implementation and stresses portability >across Smalltalks, thus not particularly Squeaky, but it works. If anyone >from Indelv is lurking, please contact me so we can do what it takes to >make it available in a manner consistent with the Indelv and Squeak >licenses. > >The toy XML parser I wrote earlier is much farther along now, and fits more >in the Squeaky way of thinking, so it might make a suitable alternative. >Hopefully I can get it out for people to play with soon. > >Regards - > >-- Duane -- David Cramer, Process Innovation Evangelist 87-1313 Border Street PBSC Computer Training Centres (an IBM company) Winnipeg MB R3H 0X4 Corporate Office Research & Development Canada

1 0

Re: xml.smalltalk.org and mod.smalltalk.org
by Peter William Lount 29 Jan '00

29 Jan '00

Hi, For some reason some people can't see http://www.xml.smalltalk.org or http://xml.smalltalk.org. Could you check each for me. Thanks. Peter William Lount peter(a)smalltalk.org http://www.smalltalk.org ---------- From: Duane Maxwell <dmaxwell(a)entrypoint.com> To: squeak(a)cs.uiuc.edu Subject: Re: xml.smalltalk.org and mod.smalltalk.org Date: January 28, 2000 2:09 PM I also think this is a wise thing to do, though I can't seem to see the xml.smalltalk.org domain to look at the details. Squeak could certainly take the lead here quite easily while the paint's still wet. Also, given the recent excellent work on encryption showing up on the list :), along with the earlier DSA support, we could build a system for "trusted" object interchange ala Perl's Penguin. An earlier pass at XML-based interchange is XML-RPC <http://www.xmlrpc.com/>, which is a very lightweight RPC mechanism. Incidentally, we've ported the "open source" InDelv XML parser <http://www.indelv.com> , but we're not quite sure what to do with it. It's based a lot on the Java/DOM implementation and stresses portability across Smalltalks, thus not particularly Squeaky, but it works. If anyone from Indelv is lurking, please contact me so we can do what it takes to make it available in a manner consistent with the Indelv and Squeak licenses. The toy XML parser I wrote earlier is much farther along now, and fits more in the Squeaky way of thinking, so it might make a suitable alternative. Hopefully I can get it out for people to play with soon. Regards - -- Duane >Hi, > >I think it's important that we have a "XML" ASCII standard for storing BOTH >"Smalltalk source code" and "Smalltalk object instances". >Interoperatability is becoming very important and being able to share >object instances with other languages and systems would revolutionize the >Smalltalk usage in the real world. > >In addition I think it's important that there be an Apache Web Server >module for Smalltalk. A ModSmalltalk. Any computer language that tightly >intergrates and interoperates with the Apache Web server in an Open Source >manner can gain prominance in the Web Site and Web Applications Market. >Smalltalk has an excellent potential to be the engine behind many powerful, >yet to be built, web sites. > >Please see http://xml.smalltalk.org and http://mod.smalltalk.org for more >of this discussion... > >All the best, > >Peter William Lount >peter(a)smalltalk.org >http://www.smalltalk.org >http://mod.smalltalk.org >http://xml.smalltalk.org > >Smalltalk.org is powered by http://www.zoku.net. =================================================== Duane Maxwell dmaxwell (at) entrypoint.com CTO http://www.entrypoint.com EntryPoint, Inc. (858)348-3040 FAX(858)348-3100 Information contained herein is my personal opinion and not necessarily that of EntryPoint. ===================================================

1 0

Re: xml.smalltalk.org and mod.smalltalk.org
by David Cramer 29 Jan '00

29 Jan '00

I'd be very interested in playing with a Squeak-based XML parser. I have been concerned for quite some time with the seeming lack of XML-related Smalltalk activity. Back in October, after just attended the XML World Conference in Ottawa, I posted a message to this mailing list regarding a conversation I had had with Dave Thomas, head of Object Technology International (OTI, now an affiliate of IBM Canada), and with over 30 years of experience in the computer industry, elected in 1997 to the IBM Academy in recognition of his pioneering work on object technology. Thomas confirmed my suspicions when he said matter-of-factly that Smalltalk was inherently superior to Java for the kind of programming needed for large-scale XML data management, but that Java had industry buy-in regardless of its immaturity and shortcomings. Unfortunately, he suggested it might take a couple of years for the realization to hit. I have a strong instinctive feeling that the Squeak community can play a pivotal role in the confluence of markup language technologies and Smalltalk programming, to everyone's benefit. I've been learning Smalltalk without a specific goal. An XML-related project in Smalltalk for me would be something I could really devote some time to. Good luck with InDelv! Regards, David At 2:09 PM -0800 1/28/00, Duane Maxwell wrote: > >Incidentally, we've ported the "open source" InDelv XML parser ><http://www.indelv.com> , but we're not quite sure what to do with it. >It's based a lot on the Java/DOM implementation and stresses portability >across Smalltalks, thus not particularly Squeaky, but it works. If anyone >from Indelv is lurking, please contact me so we can do what it takes to >make it available in a manner consistent with the Indelv and Squeak >licenses. > >The toy XML parser I wrote earlier is much farther along now, and fits more >in the Squeaky way of thinking, so it might make a suitable alternative. >Hopefully I can get it out for people to play with soon. > >Regards - > >-- Duane -- David Cramer, Process Innovation Evangelist 87-1313 Border Street PBSC Computer Training Centres (an IBM company) Winnipeg MB R3H 0X4 Corporate Office Research & Development Canada

1 0

Re: VM improvement: speeding up failing calls of functions in missing modules
by Jake Hamby 29 Jan '00

29 Jan '00

On Fri, 28 Jan 2000, Duane Maxwell wrote: > If a shared library could be closed and reopened from within the VM, you > could do some really interesting stuff under certain enlightened platforms, > such as translate, compile and test a plugin without ever leaving the > Squeak environment. Under UNIX the external stuff could be done with a > plugin that calls "system()", on the Mac this could be done with the > AppleScript plugin and CodeWarrior. Hmm, this is one of the things I was thinking about when I posted my last message about what it would take to translate simple Smalltalk or subset-of-Smalltalk programs to C using the current translator. I phrased it in terms of being able to compile simple stand-alone "Hello world" type programs to show that you wouldn't need to have the entire Squeak environment in the runtime, but I was also thinking about the implications of being able to call the C compiler in the background and reload the translated version as a module. If we could do this on the fly, that'd be an excellent compliment to something like JITter for a really high-performance VM: do quick JIT-compiles of everything, then run the C compiler in the background to do a fully optimized translation of those classes (or individual messages) that are run frequently. Hmm, I wonder if there would be any bad interactions between GC's on the Smalltalk and C side (assuming the C side was linked with something like the Boehm GC). -Jake

2 1

xml.smalltalk.org and mod.smalltalk.org
by Peter William Lount 29 Jan '00

29 Jan '00

Hi, I think it's important that we have a "XML" ASCII standard for storing BOTH "Smalltalk source code" and "Smalltalk object instances". Interoperatability is becoming very important and being able to share object instances with other languages and systems would revolutionize the Smalltalk usage in the real world. In addition I think it's important that there be an Apache Web Server module for Smalltalk. A ModSmalltalk. Any computer language that tightly intergrates and interoperates with the Apache Web server in an Open Source manner can gain prominance in the Web Site and Web Applications Market. Smalltalk has an excellent potential to be the engine behind many powerful, yet to be built, web sites. Please see http://xml.smalltalk.org and http://mod.smalltalk.org for more of this discussion... All the best, Peter William Lount peter(a)smalltalk.org http://www.smalltalk.org http://mod.smalltalk.org http://xml.smalltalk.org Smalltalk.org is powered by http://www.zoku.net.

3 2

← Newer
1
2
3
4
5
6
7
8
9
...
47
Older →

Jump to page: