Squeak-dev December 2004

squeak-dev@lists.squeakfoundation.org

194 participants
491 discussions

Re: Adding loop primitives/optimizations (was Making Set/Dictionaryetc. loops more robust)
by Joshua Scholar 02 Dec '04

02 Dec '04

Richard A. O'Keefe ok at cs.otago.ac.nz Thu Dec 2 05:11:00 CET 2004 Previous message: Accessing a particular object in GOODS causes "Space is low" warning Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] > "Joshua Scholar" <jscholar at access4less.net> wrote: > I was thinking about iteration and closures the other day. Smalltalk blocks > are becoming the equivalent of lambdas, but in Scheme, the block inside of a > loop does not have to be a lambda, and can use the outer environment I > think... So when we slow our loops down in order to create a new > environment for each iteration, we're doing something that Scheme and Lisp > didn't I think. I'll have to write some Scheme programs to be sure. > > As an old Lisper I must say I find his description of Scheme unrecognisable. > The basic loop form in Scheme is named-LET: > > (let <name> ((<v1> <e1>) ... (<vn> <en>)) <body>) > > which is precisely equivalent to > > (letrec ((<name> (lambda (<v1> ... <vn>) <body>))) > (<name> <e1> ... <en>)) > > "do" is also defined in terms of rewriting to a recursive "lambda". > "In Scheme, the block inside a loop" _is_ (the body of) "a lambda". > It "can use the outer environment" because so can all lambdas in Scheme. > And so, of course, can blocks in Smalltalk. > Scheme _does_ create a new environment (in principle) for each loop > _entry_ but not for each loop _iteration_. > And so does Smalltalk. Squeak does *NOT* "slow our loops down in order > to create a new environment for each iteration". I was looking from this example from the Squeak Swiki |blocks| blocks := (1 to: 10) collect: [:each | [each]]. blocks first value. "you would expect this to be 1, but it is 10". They were claiming that with closures the value would be 1 not 10. For it to be 1, the environment for each loop would have to be distinct from the environment from the previous iteration, just as I said. ... What they do in Scheme I'm not sure, but I have read papers (yes plural) on plenty of Scheme compilers, and most Scheme compilers analyse the code to see if they need a new environment (is any variable captured etc) and they don't bother to created a full environment unless it's needed. Joshua Scholar

1 0

Re: Adding loop primitives/optimizations (was Making Set/Dictionaryetc. loops more robust)
by Richard A. O'Keefe 02 Dec '04

02 Dec '04

"Joshua Scholar" <jscholar(a)access4less.net> wrote: I was thinking about iteration and closures the other day. Smalltalk blocks are becoming the equivalent of lambdas, but in Scheme, the block inside of a loop does not have to be a lambda, and can use the outer environment I think... So when we slow our loops down in order to create a new environment for each iteration, we're doing something that Scheme and Lisp didn't I think. I'll have to write some Scheme programs to be sure. As an old Lisper I must say I find his description of Scheme unrecognisable. The basic loop form in Scheme is named-LET: (let <name> ((<v1> <e1>) ... (<vn> <en>)) <body>) which is precisely equivalent to (letrec ((<name> (lambda (<v1> ... <vn>) <body>))) (<name> <e1> ... <en>)) "do" is also defined in terms of rewriting to a recursive "lambda". "In Scheme, the block inside a loop" _is_ (the body of) "a lambda". It "can use the outer environment" because so can all lambdas in Scheme. And so, of course, can blocks in Smalltalk. Scheme _does_ create a new environment (in principle) for each loop _entry_ but not for each loop _iteration_. And so does Smalltalk. Squeak does *NOT* "slow our loops down in order to create a new environment for each iteration".

1 0

Re: About removing global variables
by Richard A. O'Keefe 02 Dec '04

02 Dec '04

=?ISO-8859-1?Q?st=E9phane_ducasse?= <ducasse(a)iam.unibe.ch> wrote: by global I meant variable holding something else than a class and stored in Smalltalk. That is not what 'global' normally means. SystemOrganization is one that is badly accessed and should not since it should be accessed via Smalltalk organization. There are a number of odd things about SystemOrganization. One is that its class comment says v My instanceS provide an organization for the classes in the system... ^ and yet there is only one instance, and as far as I can tell, only ever meant to be one instance. If, instead of being an(the) instance of SystemOrganizer, it were a subclass of Object holding an(the) instance of SystemOrganizer as a class variable and with all the methods of SystemOrganizer mirrored as class methods of SystemOrganization, it would, for all practical purposes, be indistinguishable from what it is now, yet by the definition of 'global variable' above would no longer be a 'global variable'. I can't think of anything I'd do with SystemOrganization that I wouldn't normally do through the UI instead, so it's no skin off my nose if it's put in Smalltalk. I do wonder how one decides where to put it, though. It has never been clear to me just how much must be shared between projects and how much can be different. Can different projects have different organisations, and if not, why not? > Now, if we had 1800 global named constants (other than class names), > then "globality" _would_ be a problem, and we'd need to partition that > name space somehow. But with just 12, let's focus on the right > problem. > Having 12 globally visible named constants is _not_ a problem. it is your point of view. An extra concept that does not seems to be needed You seem to have completely missed my point. My point is that Squeak ALREADY HAS the required concept (global named read-only binding). I am not proposing a new concept, not in the least. My proposal is simply that we use an *EXISTING* implemented concept in Squeak to enforce what is already the intended use of a number of globals (like 'Smalltalk' itself).

1 0

[ENH][FIX] PreferencesFlagshipTypo
by humasect＠shaw.ca 02 Dec '04

02 Dec '04

"Change Set: PreferencesFlagshipTypo Date: 1 December 2004 Author: Lyndon Tremblay Corrects typo in Preferences>>isFlagship"

1 0

3.6-3 VM on Sparc NetBSD 1.6.2
by Bruce ONeel 02 Dec '04

02 Dec '04

Hi, Continuing my quest to run as many odd and outdated systems as possible, I've now managed to scare up a sparcstation 4 for the home network. Oddly, given that 3.6-3 runs on i386, ppc, and mac68k versions of NetBSD it crashes reading the sources file on startup with a bus error on SPARC. It happens on the ftell at line 172 of sqFilePluginBasicPrims.c and yet both f and f->file seem to point to reasonable things. This is with gcc 2.9x or so. Is there something obvious I've missed? The last VM version which does seem to run is 2.8pre3. I'm a bit confused here. Thanks! cheers bruce

3 2

Re: "A Word of Caution" to all newcomers
by Hannes Hirzel 01 Dec '04

01 Dec '04

John Pierce wrote: > ------------ > | A note to students, parents, and teachers | > ------------ > If you are a student, parent, or teacher looking for well-documented > and easy-to-use educational software, you may want to visit > SqueakLand.org, where you can find some great educational projects > built in Squeak, but delivered in an easier-to-use distribution of > Squeak. > > If you are still eager to explore this base distribution of Squeak, > there is lots of help available in the form of web-accessible > tutorials and documentation and on-line via the Squeak mailing list > from numerous knowledgeable users of Squeak. Access to the mailing > list can be found at: > http://squeak.org/documentation/index.html#mailing. > ------------ Excellent! This is basically the same message but put in a much friendlier way. I would like to see _this_ note included in the 3.8 or 3.9 image instead of the 'A word of caution' note. Hannes

1 0

State of Squeak CVS repository at SqueakFoundation
by Alexander Lazarević 01 Dec '04

01 Dec '04

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 laza@eslapion:/var/tmp$ cvs -z3 -d ':pserver:anonymous@cvs.squeakfoundation.org:/squeak' co . cvs server: Updating . cvs server: failed to create lock directory for `/squeak' (/squeak/#cvs.lock): Permission denied cvs server: failed to obtain dir lock in repository `/squeak' cvs [server aborted]: read lock failed - giving up ? Alex -----BEGIN PGP SIGNATURE----- Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iD8DBQFBrZPeYiF2wSTEZ9gRAlgUAKCeWuuhWqOh4aCvvqiSoKXTxUSiuwCgySKT 4RfXPAkRhuvJzqIDa9nhlFo= =65la -----END PGP SIGNATURE-----

3 3

[BUG][FIX]numbers in scientific notation with capital E are not handled correctly
by Baveco, Hans 01 Dec '04

01 Dec '04

I have often been bit by this feature of squeak. '1234.123E-5' asNumber produces 1234.123 because the exponent part is ignored. Some numeric software produces output in this format, and reading datafiles from such programs may lead to very hard to trace anomalies (especially when scientific notation is only used for some of the numbers). A simple one-liner change in Number>>readRemainderOf:from:base:withSign: seems sufficient to handle capital E, see attached fileout. I changed "peekChar = $e | (peekChar = $d) | (peekChar = $q)" to "((((peekChar == $e) or: [peekChar == $E]) or: [peekChar == $d]) or: [peekChar == $q])" (I compensated for the extra check by replacing the bars by blocks). I can't imagine there are any unwelcome side-effects associated with this change? ==================================== Hans Baveco <<Number class-readRemainderOffrombasewithSign.st>>

2 1

Mac OSX display depth colour oddness
by Peace Jerome 01 Dec '04

01 Dec '04

Mon Nov 29 21:15:38 CET 2004 John wrote: >So what's the problem with 3.7.5bx? My experience is that the problem I've seen is in all vms I've used and 3.7.5b gives similar results to 3.7.4b. >Also I'm not sure what you mean by your comment that earlier squeaks (image:? VM?) wouldn't recognize a 32bit screen depth? That was an old recollection and impression I had from my early experiences with squeak. I just went back and tried some old vms with an old image and it did not seem to be true at least with 3.2.8b. Maybe I was mistaken. I also tried to answer your questions by trying to find a way to pin things down. With vm 3.7.5b classic and image 3.9a-6461 I did the following: in a fresh project. Grab a ruler from objects. Made it 512@32 with grow handle With color properties set a ramp white-->black in direction 512@0. Then I played with combinations of squeak screen depth from the appearance menu and hardware screen depth. from the apple control panel (actual the retractable tab version but you get the idea.) When the squeak screen depth is set to 32 and the apple is set to thousands of colors the ruler showed 32 steps of gray. When the squeak screen depth was reduced to 16 the ramp appeared smooth no steps. When the squeak screen depth was put back to 32 and the apple set to millions of colors the ramp looked smooth. When the screen depth now was reduced to 16 colors the ramp looked smooth. I did grab patches on the various combinations and noticed some interesting things happened when I fooled around with squeak and apple screen depths. But there were some problems with reproducibility and analysis of the results. The obvious work around is to have the screen depth and hardware depth match as often as possible. While I believe there is a findable bug in all this (though maybe not in the vm) I don't think its a good thing for me to give a high priority to. If I stumble over something that illuminates it further I may get back to it. If not, not. If I find something solid I'll post it on mantis. Thanks for your response. Yours in service --Jerome Peace __________________________________ Do you Yahoo!? Meet the all-new My Yahoo! - Try it today! http://my.yahoo.com

2 1

DB-oreinted application
by Blake 01 Dec '04

01 Dec '04

OK, Squeakers, I'm in a situation where I can see myself being poised to suggest a language tool for an upcoming iteration of a retail product. It's a ways off, but it is, I think, inevitable. We are currently using an Access-like tool to develop the application, which has around 200 different tables and screens. It also has many variants. From a design standpoint, it cries out for object-orientation, but of course, we have bupkus as far as genuine OO tools at hand. I'm considering suggesting Squeak (or any Smalltalk) as the platform but there are a few issues, some more easily resolvable than others. I'd like to hear some ideas on this: -> The lure of the Access-like tools is the fact that any moron can design a table or set and then create a form or grid based on said table. Forms generally need a lot of massaging and re-formatting so they're not a huge deal, but I've seen no grids. Having built a few grids in my day I'm not keen on building more. -> We'll have at least 4-5 people working on the same application. One of the nightmares of the current environment is that it was not designed for collaborative work. (I confess also that STers disdain for files is not one I understand. Working in an image makes some things very convenient and other things very inconvenient.) I've read a lot of caveats regarding Monticello and I'm not quite sure how well it would work. Anyone with experience here? How will Croquet factor in, given that it's all about collaboration? (I can see certain elements of Croquet being very useful for working out bugs and brainstorming, but I don't know about source control.) -> Right now the bosses are concerned with the fact that the app doesn't look like a "normal Windows app" that people are used to. I'm aware of the problems and solutions to this in Squeak, but I also suspect this will be a non-issue shortly. MS seems to have abandoned Win32. -> The web-based squeak stuff--not Seaside, but the plug-in--makes it appear as though any squeak application can be put on the web with a flick of a switch (I suppose especially if the data storage is handled correctly). I am somewhat concerned about interoperability between Squeak and .NET/J2EE Thoughts? Thank you, kind folk, in advance. ===Blake===

6 8

Jump to page: