Squeak-dev January 2000

squeak-dev@lists.squeakfoundation.org

131 participants
468 discussions

Re: SqueakV2.sources
by Bruce ONeel 31 Jan '00

31 Jan '00

Hi, There are two .sea files, one straight from Squeak Central which doesn't include the sources file and the other named _mac.sea which does. By default all ofthe other bundles include the .sources file even though that makes the download longer. cheers bruce Bob Arning <arning(a)charm.net> wrote: > On Sun, 30 Jan 2000 18:57:20 +0530 "Mayuresh A Kathe" <mayuresh(a)vsnl.com> wrote: > >Everytime I start Squeak or save any changes I get a message specifying that > >Squeak is unable to find SqueakV2.sources file. I have downloaded the .sea > >file and not as seperate files. > > > >What to do, where can I get the file from?? > > Mayuresh, > > Since the .sources file is the same from one minor release to the next, it is not usually included in the bundle of files that comprise that release. You can probably find it lots of places. Here's one: > > ftp://st.cs.uiuc.edu/pub/Smalltalk/Squeak/2.0/SqueakV2.sources.sea.bin > > Cheers, > Bob

1 0

[ENH][GOOODIE] InfoAgent (Scamper enhancements)
by thomas.mahler＠itellium.com 31 Jan '00

31 Jan '00

Hi all, I'd like to announce the release 1.0 of InfoAgent >From the preamble: InfoAgent enhances Scamper with - bookmarks (incl. sharing bookmark-file with Netscape Navigator) - convenient ui for using web-based search engines - tracking changes in web-pages See the project's homepage at http://minnow.cc.gatech.edu/squeak.InfoAgent for more details and latest downloads enjoy, Thomas

1 0

Re: Class comments within .change files are anonymous doits
by R. A. Harmon 31 Jan '00

31 Jan '00

At 06:10 PM 1/30/00 -0600, Les Tyrrell wrote: >Lex Spoon wrote: >> >> One easy way to fix this would be to use ANSI SIF as the standard >> interchange format, instead of traditional fileouts. SIF makes >> everything quite explicit. Someone has posted SIF code to this list >> before; I don't recall the details. [snip] > >One thing that could interfere, though- I don't know if the >standard for SIF will accomodate Squeak's style and stamp >information- the style information often contains hyperlinks, >and the stamps might contain versioning information. [snip] In a previous SIF version I had accommodated style information in annotation comments. I put the comment out in a comment annotation and the style as a run array store string value in a second annotation with a key of "commentStyle". This mechanism was general enough to handle style in method source, etc. I also added a similar mechanism to handle multiple categories with the primary category in the "category" annotation and an array of secondary categories store string in a second annotation with key "categoryList". I don't think this broke the standard, but it added some complexity so I dropped it when I ported my Squeak SIF to Dolphin. I've still got the code to do it. It was an experiment in how to do it within SIF limits. I think a portable style standard should probably be used. -- Richard A. Harmon "The only good zombie is a dead zombie" harmonra(a)webname.com E. G. McCarthy Spencer, Iowa

1 0

Re: Full Closures - when do they close?
by Robert Withers 31 Jan '00

31 Jan '00

yes it is close isn't it. ;-) I had no idea about #fixTemps. This second example has me in knots again. I would think that foo would have seperate bindings, in a full closure system, in the two different blocks. Stephan Rudlof wrote: > > Vassily, > > thank you. You have a deep understanding of this - Smalltalk version > specific - stuff. [snip] > Just to check understanding of binding levels: > > Method > | foo | > ^Array > with: [:arg | foo := arg] > with: [foo] > > Squeak BlockContexts: - 'foo' is on the same level as 'arg', > Real BlockClosures: - 'foo' is on another level as 'arg'. The real BlockClosures example here is the one that threw me. I would think there would be three bindings of foo here and only two bindings when the Array is returned. So let me construct an example: Method |array foo| foo := 10. array := Array with: [:arg | foo := arg] with: nil. foo := 20. array at: 2 put: [foo]. ^array "Before the return..." {foo}-->20 {array}--> [1, 2] {array at: 1}--> {block} | {foo}--> 10 "arg doesn't have a binding yet {array at: 2}--> {block} | {foo}--> 20 Is this not our picture? Does it have anything to do with foo on the lhs in the first block? thanks, Rob [snip] > > > Is it correct to say > > > aBlockContext fixTemps > > > in Squeak behaves like a BlockClosure in VW? > > > > Great question! > > > > No. One of the examples I gave in this thread: > > > > | foo | > > ^Array > > with: [:arg | foo := arg] > > with: [foo] > > [snip] > > "i" and the example works. #fixTemps copies a little too much though--the > > whole environment frame of the containing method, including variables such > > as "foo" in the second example. If you add #fixTemps to the example above, > > blocks would use separate bindings for "foo". Because closures do not copy > > bindings (at least conceptually), the behaviour is different. > >

2 1

Re: Full Closures - when do they close?
by demiourgos＠smalltalk.org 31 Jan '00

31 Jan '00

On Sat, 29 January 2000, Robert Withers wrote: > > This raises a question, Jan. What is the full set of variable bindings > in Smalltalk (VW). Global, makes some sense, but what about semi-global > and local. Would this be the difference between external bindings in a > closure versus argument bindings to the closure? > > The next step from here seems to be the impact of namespaces on these > concepts. Or rather, do these concepts help in the structuring of a > namespace? [snip] Well, before I embraced Smalltalk, I was enthusiastic about functional languages, ones where functions were first class objects. The power there came from two places. One was the ability to map things like [:listItem | ...] over datastructures. In some sense the mapping was "parameterized" to accept the block as its "innards". The second was the emerging calculus of "computing" or "deriving" programs from specifications. That's pretty erudite. But, while managing huge states (like a global COMMON) was surely something to be avoided, the need for state always pressed when one was programming with pure functions. Indeed, SCHEME was an attempt to handle the needed state stuff elegantly. But objects are better, acknowledging that what's needed are self-contained packages of code+state with well-defined interfaces. Thus, to attempt the closure idea within Smalltalk or Squeak, one realization might be to package into the closure a copy of the objects that are visible at the moment the closure is formed. On a monolithic processor, these all could be by reference, and one could have some kind of epoch label or tag to identify the levels of values, differentiating between values held by different instances or between values held by different classed. The idea would be to build the realization so that variables were like base registers in LISP, and objects would "hang" off of them like property values, indexed by these tags. If one distributes this across a network, then the pointer model either needs to be enriched to object pointers, as in the XML framework, or copies need to be sent. I think I agree with Vassily that the built in Smalltalk mechanisms for blocks and processes are a superior way to go. Keeping track of closures tucked away in variables is pretty hard when there are very many. Indeed, most programs which use closures make and use one or two in very special and restricted places. I think it's rather like exception handling in most languages. One can use that as one's primary control structure, but that doesn't make it a good idea. Hope I was clear. --jtg ______________________________________________________ Jan Theodore Galkowski www.smalltalk.org/ http://www.scguild.com/usr/1707I.html demiourgos(a)smalltalk.org www.marssociety.org/ ********************************************** PGP Key Fingerprint: 2757 F86D AA51 677D 38D7 964B 9A8D 7852 A494 3790 ********************************************** Get my Public Key from my home page at: http://home.stny.rr.com/algebraist/ ********************************************** ______________________________________________ Get free e-mail at http://www.britannica.com

2 1

[JOB] w/ Randy Smith at Sun Labs
by John.Maloney＠disney.com 31 Jan '00

31 Jan '00

Dear Squeakers: Randy Smith is looking for someone to help him in an interesting project at Sun Labs. For those of you who don't know Randy, he was one of the co-designers of the Self language and also the true guiding light behind the design of Morphic. (Randy and I built the first version of Morphic in Self, before I left Sun Labs to join Alan Kay's group at Apple). Randy has done many other very cool systems, including the Alternate Reality Kit (written in Smalltalk at Xerox PARC), experiments with using audio in the user interface, and experiments with collaborative problem solving using shared online spaces (one system was based on the Alternate Reality Kit, the other based on a multi-user version of Morphic.) Randy used to be a Physics professor, and thus he brings a physicist's perspective to every project that he works on--a desire to boil things down to a small number of powerful principles that can be applied universally. Randy is also a terrific mentor and very easy to work for and with. I don't know much about this latest project, but I'm sure it will be fascinating, and this is a chance to get in at the very beginning. And, although you will apparently have to program in that vastly inferior language named after a caffeinated beverage, there are any number of interesting ex-Self-group folks around to commiserate with... If you are interested, please correspond directly to Randy at the address below. -- John -------------------------------------------------------------------- >From Randy Smith: Position in Sun Microsystems Research Labs I anticipate hiring someone to work on a data visualization framework for Java. My hope would be to find a fine programmer who is familiar with UI frameworks and/or visualization techniques. (Master's or PhD in Computer Science desired, but not absolutely required.) The goal is to create a kit in which multi-dimensional data is represented and dynamically filtered in both conventional and innovative ways (such as fisheye-like zooming techniques in spatial and non-spatial dimensions such as color and transparency. Perhaps 3D will play a role.) Ultimately, the end-user should be able to explore, playing with the way the data is mapped to the display. Creative ideas here will be welcomed. The position will be in the San Francisco Bay Area (Palo Alto / Mt. View), reporting to me (Randy Smith). If interested, please send a resume or resume-like e-mail note to randall.smith(a)sun.com. --------------------------------------------------------------------

1 0

Re: Full Closures - when do they close?
by Jesse Welton 31 Jan '00

31 Jan '00

Robert Withers wrote: > > Are there any special rules for the internal state of an external value > binding? My guess is that if you don't change the value of the outer > reference, yet you modify the internal state of that object, then the > closure binding to that value will refer to the object which had it's > internal state modified. Is this the case? Yes. The bindings just hold onto the object references, so internal state is shared as in any case where two variables are bound to the same object reference. -Jesse

2 1

Re: VM improvement: speeding up failing calls of functions in missing modules
by Stephan Rudlof 31 Jan '00

31 Jan '00

"Raab, Andreas" wrote: > > Stephan, > > > I didn't know that module names are always stored as symbols, and in > > dubio I try to fulfill the specification of the primitive. > > Is it safe to expect no changes regarding this point? > > Yes. I can't see a reason for having these names not being symbols. Note > that in the eventual case that there is a String stored in the module name > the only thing going wrong would be that you'd store the string explicitly > in the missing module cache (no problem as long as not all modules are > strings which they really shouldn't). I see a problem here: In Interpreter>>isInMissingModules: there I have removed the string/symbol compare for a merely oop compare: So different in oops, but equal in bytes strings would fill up the cache fast, if many different functions from the same module - stored as strings with different oops - are called. This could lead to a bug: there is no caching anymore, if some (> cachesize) different functions each stemming from another module are called in sequence! > > Andreas > > PS. From the v8 change set: Now that the missing module cache is actually in > the special objects we could completely avoid clearing it from the VM level > - this could be done during startup from the ST side (or whenever we feel a > flush is appropriate). But in either case it has to be cleared during startup time. What do you think of is a good time? The cache has to be cleared early before _every_ possible module call. Stephan

1 0

Re: [ANN] distribution of Squeak programs
by Stan Heckman 31 Jan '00

31 Jan '00

I have a suggestion for the .tgz file. I think users of this format expect a readme. It might contain the following. To start, run the executable "hanoi" from the directory it and "squeak.image" are found in. If you have administrator privilege and wish to install hanoi for all users on an FHS compliant system, copy "hanoi" and "squeak.image" to /usr/lib/hanoi, and "hanoi.sh" to /usr/bin. Generating the above text automatically may be just as much work as automatically generating a file named install.sh containing #!/bin/sh # install for all users in /usr/bin and /usr/lib/hanoi # requires administrator privilege mkdir /usr/lib/hanoi cp hanoi squeak.image /usr/lib/hanoi cp hanoi.sh /usr/bin and a readme containing To start, run the executable "hanoi" from the directory it and "squeak.image" are found in. If you have administrator privilege and wish to install hanoi for all users, install.sh might do what you want. -- Stan

1 0

Re: VM improvement: speeding up failing calls of functions inmissing modules
by Stephan Rudlof 31 Jan '00

31 Jan '00

Duane Maxwell wrote: > > Actually, while you're mucking around in there, I'd kinda like the ability > to make the VM close external plugins in order to make the debugging on > them easier. At the moment, to test a new version of a plugin, we have to > quit and restart the VM. It doesn't seem to be so easy to get rid of an already installed shared lib on Linux (don't know for other platforms): If I'm removing a plugin module while Squeak is running, a previously called function from this module is further available though. A file 'sync' doesn't change this behavior. Therefore it seems to be necessary to browse all calls to e.g. a module like 'LargeIntegers' by searching for methods containing lines like <primitive: 'primDigitAddWith' module:'LargeIntegers'> by Smalltalk browseMethodsWithSourceString: '''LargeIntegers''' , and then? Currently I don't know, how the cache mechanism for the function pointer works, which is also given as a parameter (in an array) to Interpreter>>primitiveExternalCall (look there)... It would be necessary to clear all these pointers in the whereever cache: then they would be recomputed automatically with next call of the corresponding function in a renewed module (at least I suppose this for Linux). > > 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. This would be very neat. I think one challenge is to ensure same semantics on all platforms... 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

1 0

Jump to page: