Squeak-dev May 2003

squeak-dev@lists.squeakfoundation.org

204 participants
710 discussions

Re: KCP feedback
by Daniel Vainsencher 05 May '03

05 May '03

You kept stuff small, you kept changes logically grouped, and you made your intentions clear in the change set comments, I liked that :-) I look forward to more improvements and cleanups to the Kernel. And keep bringing up those discussions like about breaking up SystemDictionary, they are very important. Lets us know where you're going... Daniel Stephane Ducasse <ducasse(a)iam.unibe.ch> wrote: > Hi daniel > > Thanks ***a lot*** we will fix what you found today I hope. > You went quite fast. I like that. > > Stef > > > On Monday, May 5, 2003, at 01:05 AM, Daniel Vainsencher wrote: > > > Ok, I've gone through the KCP stuff. I know we orginally decided to do > > the removals first, but I had an urge, and this doesn't touch so many > > different classes that I'm very worried about clashes. Anyway, it seems > > as the designated reviewer for this specific project I got back from my > > vacation before the removals were done, and I don't see much reason not > > to merge these now, but I don't mind checking the conflicts if this > > proposal isn't accepted. > > > > I've reviewed almost all the code changes, but didn't test extensively. > > I did find one bug described below, but we can easily fix small bugs as > > we find them, I'm mostly reviewing the design decision, which seems > > very > > sound. In a few places a little more intent is warranted, because we > > will actually want people to continue in the new direction, so I asked > > questions below. > > > > Anyway, everything currently posted (and enumerated explicitly below, > > since web pages change), is approved. > > > > Loads of thanks to the KCP team - hopefully we will soon see the kernel > > just be a kernel. > > > > Bug - ctrl-w doesn't work because > > SystemNavigator>>allSelectorsWithAnyImplementorsIn: tries to get a > > system navigator. Maybe a quick review of all the query methods is in > > order, to see they don't have such stuff. > > > > 0011: > > Behavior>>becomeCompact has the line (also a similar method) > > cct _ self environment compactClassesArray. > > Since you're taking SystemDictionary apart (woohoo!) be careful to then > > also fix all these references you change to self environment. I'm not > > sure this isn't making things more difficult for you, but anyway, you > > can later look for environment senders, I guess. > > > > About SystemNavigation - it has a nice comment that does answer some of > > the questions I had, but can you say something about how it's supposed > > to fit into everything? for example, why all the similar > > implementations > > of self systemNavigator and not one implementation in Object (wait! > > don't shoot!) in a class extension? it's not very clear in what case > > you'd need state for this bunch of utility methods, so what examples > > did > > you have in mind when you made them instance side? > > > > With all the deprecation going on, is there a test that finds all > > deprecated methods, finds all their senders, and makes sure it's 0, or > > that the senders are to self environment or self navigator? seems to me > > safer than all the "this should be deprecated" tests, because one of > > these tests might be missing. It seems a bit more complicated than > > this, > > because some are not renamed, just moved, but I think some partial > > sanity checks of this type might be good. > > > > 13, 14, 16, 18, 19, 21: approved. Thank you for moving to the > > self-sufficient-changeset style (where deprecation of a method and it's > > redefinition in SystemNavigation are in the same changeset). It's much > > easier to check. > > > > 22, 24, 26, 28, 30, 32, 34, 38, 40, 41, 43, 45, 46, 48 50 52, 55 and 57 > > approved. > > > > Daniel > > > > > Prof. Dr. Stephane DUCASSE [ | ] > http://www.iam.unibe.ch/~ducasse/ > "if you knew today was your last day on earth, what would you > do different? ... especially if, by doing something different, > today might not be your last day on earth" Calvin&Hobbes > > Open Source Smalltalks: http://www.squeak.org, > http://www.gnu.org/software/smalltalk/smalltalk.html > Free books for Universities at > http://www.esug.org/sponsoring/promotionProgram.html > Online Free Books at > http://www.iam.unibe.ch/~ducasse/WebPages/FreeBooks.html

1 0

Let's awake!
by diegogomezdeck＠consultar.com 05 May '03

05 May '03

Hi guys... The following is an email I'm thinking to write for weeks. To say it shortly: Squeak is dieing in front of us... Since SqC leaves the "dictatorship" role we don't find the way to continue. Despite the apparent organization and procedures we had created, simply in the last months we was not able to satisfy the expectations. To probe my point just take a look to the amount of fixes waiting for approval and the changes produced by MCP and KCP. Personally, I paused my work in MCP until something happens in our community. I invest most of my free time in Squeak and I don't feel this sacrifice is producing results. >From my point of view We're trying to organize so much, and in the reality we have pages and pages describing a process (the harvesting process) that don't work. The bureaucracy is killing us. The steps to get a fix approved are, in most of the cases, more difficult and time consumer than the fix itself. We have lists of people with roles (guides, harvesters, etc) but in the reality only a few of them are active and working. We have big problems and I don't feel the community realize of it. Cheers, Diego Gomez Deck PS: I'm talking for my own and my thoughts are not the MCP-team opinions.

6 10

Squeak: the Open-Source Smalltalk, Evaluation for Teaching
by Stephane Ducasse 05 May '03

05 May '03

Hi all the 27 of May I will do a presentation at the University of Geneva (http://osiris.unige.ch:8899/sem/ns?a=sem_present) on Squeak in the context of teaching OOP. I hope to have a presentation ready soon for feedback :). If you have any idea just send me emails. I already have a lot to say but By the way what is the best way to have a good looking fonts (because the default way I was doing the slide using trueTypeMorph edit contents...is killing me to the point I was thinking to prototype the talk in PPt before doing it in Squeak). Stef Stef www.iam.unibe.ch/~ducasse/ [ | ]

4 3

Re: ReadWriteStream issues
by Anthony Adachi 05 May '03

05 May '03

goran.hultgren@b... wrote: >However, if you try your code with ReadStream instead it works. Okay, I tried ReadStream and it works up to a certain point with the "Squeak: A Quick Trip to ObjectLand" book's examples. For instance these messages work as expected. s :=ReadStream on: #(1 2). s next. s next. s atEnd. s reset. s next. s position. However, since the class is ReadStream I assume that means it isn't meant to support writes? The above and following examples from the book used ReadWriteStream. If these sample message sends are sent to an ReadStream instance instead they produce errors. Which, I assume comes at no surprise. s nextPut: 7 "Error: this message is not appropriate for this object." s nextPut: 8 "ditto" s nextPutAll: 'AB'. "ditto" > I can't remember exactly but I think it was something about the > inherited class side #on: method that evidently got "inherited" a bit by > mistake. I think the Stream and Collection hierarchy are full of those > "mistakenly" inherited little methods. Does this mean one should use a different class other than ReadWriteStream if one needs the same state and behaviour as shown in the book? Or does one need to call a method other than on: or with: ? > I actually even hear a small bell ringing somewhere back in my head that > I posted about this issue... or at least something related a loooong way > back. I think I did anyway, but it was so long ago I couldn't find any > reference to it. Yesterday, I searched the list for "ReadWriteStream" but I didn't find much. I just did a search now for "ReadStream" and didn't see anything about this issue either. Does anyone know if there been a change to the ReadWriteStream class in Squeak since the book was written? Or if this might be a case of a printing error in the book and the wrong class was shown in the examples? Thanks, Anthony __________________________________ Do you Yahoo!? The New Yahoo! Search - Faster. Easier. Bingo. http://search.yahoo.com

1 0

Re: Context Stack Speedup
by Anthony Hannan 05 May '03

05 May '03

Hi Bryce, sorry for my late reply. Bryce Kampjes <bryce(a)kampjes.demon.co.uk> wrote: > I like your design a lot. It really reminds me of optimized contexts > and dynamic deoptimisation needed for adaptive compilation. > > Anthony Hannan writes: > > Let me say that this design adds a new optional kernel class to > > the image making images that use it incompatible with older VMs, but > > the new VM will still be able to run older images and load old projects > > (image segments). So it is basically a backwards-compatible image > > format change. Let me take this back. We can make this enhancement completely backwards and forwards compatible by converting all ContextStacks back to context objects before saving the image. > > New Context Class > > > > A ContextStack is a sequence of method contexts embedded in its > > indexable fields (its stack). ContextStacks, MethodContexts, and > > BlockContexts are chained together forming a full execution stack. When > > the top context within a ContextStack is accessed it is popped out into > > its own MethodContext and kept in the chain at its original position. > > > > For example, suppose context A is a suspended context with B, C and D as > > its senders in that order, and B, C and D all resides within the same > > ContextStack Z. Sending #sender to A will cause B to be popped out into > > its own context and Z would only have C and D remaining in it. If we > > continue sending #sender until we reach the end, all will be in their > > own context and no ContextStack will remain. > > > > So anytime we need an object reference to a context it is separated out > > into its own context. So, the debugger, etc. will never see > > ContextStacks since their frames are converted to contexts as they are > > accessed (through #sender). > > > > The VM will maintain contexts inside a single ContextStack, unless it > > executes pushThisContext then it will separate out the current context > > into its own context and start a new ContextStack after it. Block > > contexts and home contexts of blocks that return to it (^) also > > execute pushThisContext so they will have their own contexts as well. > > How does the VM know what contexts require their own contexts? Those that call thisContext. After that #sender in the image will extract each sender out when called. > Are you adding a flag to MethodContext? For Squid I am adding a flag to MethodContext. But for the existing image format I would wait until pushThisContext bytecode and then extract out the context if not already done so. > For block creation separate bytecodes could work but a send doesn't > know what method will be sent. Right, it won't know until pushActiveContextBytecode is executed, then it would retroactively pop the current context into its own MethodContext. > I'm interested for a couple of reasons. One, the design looks very > similar to inlined contexts in a system with dynamic compilation. The > debugger interface screams dynamic deoptimisation to me. Two, to keep > things simple for those of us messing around in the interpreter it > would be nice to disable the ContextStacks. Three the incrementalist > in me really likes the idea of being able to switch off the optimization > selectively if it's buggy, though that probably isn't necessary here. > > How different does the ContextStack look to the interpreter? Or how > widespread are the changes? Are they all isolated in the send and > return bytecodes or do all the stack bytecodes need to be modified? Only the send and return bytecodes would be modified. Some tricks would have to be played with variables like localHomeContext to make the rest of the bytecodes work with both stacks and context objects. Like the localHomeContext would point to 6 slot under the stack frame since localHomeContext assumes a MethodContext or BlockContext object with 6 instance variables. > Is it possible for a ContextStack to ever end up in a debugger? No. > Say if a context switch happens when execution is in a ContextStack then > a debugger is placed on the suspended process. Am I missing something? Along with ContextPart>>sender, Process>>suspendedProcess (which the debugger calls) will also extract out the context object if it is pointing to a context stack. > What was slowing down block closures? Are there a few common cases that > could be easily optimized? Blocks that don't access surrounding closures > come to mind. The main slow down for closures is the creation of new context/activation for each invocation. BlockContexts reused a single context/activation making it faster. If we moved activations to a stack then this would speedup all activations including closure activations. I am not working on ContextStacks for the current Squeak VM. Instead I am working on them for the new Squid VM. But maybe someone will want to implement them for current Squeak. I think the idea is pretty straight forward. If not I will be happy to clarify. Cheers, Anthony

1 0

[FIX] browsingCodeFiles
by Magistrello Alejandro (SFA) 05 May '03

05 May '03

hi all, two fixes, 1 - when you want to open a changeList browser for a file that is readOnly you a get a nil dNU: browseStream: because the file wasn't opened in write mode, this is fixed by opening the file in readOnlyMode. 2 - if you open a code-file browser on a File that contains the definition of a non-existing class in the system and goes from the classSide to the instanceSide you'll get: PseudoMetaclass/PseudoClass dNU: #compiledMethodAt:ifAbsent: this one is fixed by testing (aClass isKindOf: PseudoClass) in CodeHolder.didCodeChangeElsewhere, as one can't change code on a FileContents browser just answer false one file that gaves this error is <http://minnow.cc.gatech.edu/squeak/uploads/3030/AppRegistry.cs> regards Alejandro <<browsingCodeFixes-asm.1.cs.gz>>

1 0

ReadWriteStream issues
by Anthony Adachi 05 May '03

05 May '03

Hello I've been reading the "Squeak: A Quick Trip to ObjectLand" book and have come across some problems with getting the examples using ReadWriteStream to work (on page 174-175). I don't get the same results as the book states as one should expect. For instance, the following is a series of message sends from the book and what it says are the expected returned values: s := ReadWriteStream on: #(1 2). s next. "expected returned value should be 1" s next. "...should be 1" s atEnd. "...should be true" s reset. s next. "...should be 1" s position. "...should be 0" However, this is what I actually got: s := ReadWriteStream on: #(1 2). s next. ".result is nil while expected result should be 1 according to the book" s atEnd. "results in true" s position. "Answers 0" s reset. s next. "result is nil should be 1 according to the book" s position. "Again, answers 0 while it should be 1" s atEnd. "true again despite reset message send" s contents. "#()" I stepped through the on: message sent to ReadWriteStream and noticed that the read limit is set to 0. I tried sending the with: message to see what happened. s := ReadWriteStream with: #(1 2). s next. "result returned is nil" s position. "2" s atEnd. "true" Again, the next message returned nil instead of 1. Although, this time the position messaged returned 2. The results are closer to what the book expects if a reset message is sent and the with: creation method is used instead of on:. s := ReadWriteStream with: #(1 2). s reset. s next. "1 as expected" s next. "2 likewise, as expected" s atEnd. "true" Has Squeak's ReadWriteStream class's interface changed since this book has been published or something other problem? I'm using Squeak 3.5 on Mac OS 9.1 Thanks, Anthony __________________________________ Do you Yahoo!? The New Yahoo! Search - Faster. Easier. Bingo. http://search.yahoo.com

2 1

RE: Squid VM
by Jarvis, Robert P. (Contingent) 05 May '03

05 May '03

If I understand correctly, the Self VM was fast because a whole lot of work went into native code generation technology. See http://www.sunlabs.com/nova/cgi-bin/index.cgi?a=r&s=9343&e=9356&f=%2Fhtdocs% 2Fself%2Flanguage.html&i= Bob Jarvis Compuware @ Timken > -----Original Message----- > From: Niko Schwarz [mailto:niko.schwarz@gmx.net] > Sent: Monday, May 05, 2003 8:24 AM > To: The general-purpose Squeak developers list > Subject: Re: Squid VM > > > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Am Sonntag, 4. Mai 2003 23:28 schrieb Michael van der Gulik: > > > - And would it be possible to make a language-independant > VM, perhaps > > with modifications to respective compilers, that could compile and > > interpret Squeak, ANSI Smalltalk or Squid. Redesigning a > wheel is better > > than reinventing it. > > yes, the idea is tempting. A while ago I asked myself how > difficult it would > be to create a Self environment, and run squeak images in it. > I mean: the Self VM is said to be spectacularly fast, and > Self has all the > advantages that Anthony wanted, and most of all: the Squeak > stuff would not > be lost. > > regards, > > nick > - -- > Molecule, n.: > The ultimate, indivisible unit of matter. It is distinguished > from the corpuscle, also the ultimate, indivisible unit of > matter, by a > closer resemblance to the atom, also the ultimate, indivisible unit of > matter ... The ion differs from the molecule, the corpuscle and the > atom in that it is an ion ... > -- Ambrose Bierce, "The Devil's Dictionary" > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.2.1 (GNU/Linux) > > iD8DBQE+tlfl3P3nmXTVnC8RAv3aAKD5dF0adqbIbXeyn0BenxY55Z0NTgCeMg2o > RZImjxqKvKLbUFQ08joJF0U= > =Jnt7 > -----END PGP SIGNATURE----- > > ********************************************************************** This message and any attachments are intended for the individual or entity named above. If you are not the intended recipient, please do not forward, copy, print, use or disclose this communication to others; also please notify the sender by replying to this message, and then delete it from your system. The Timken Company **********************************************************************

3 2

RE: Squid VM
by Jarvis, Robert P. (Contingent) 05 May '03

05 May '03

Try this: http://research.sun.com/research/self/ Bob Jarvis Compuware @ Timken > -----Original Message----- > From: Niko Schwarz [mailto:niko.schwarz@gmx.net] > Sent: Monday, May 05, 2003 8:35 AM > To: The general-purpose Squeak developers list > Subject: Re: Squid VM > > > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Am Montag, 5. Mai 2003 14:32 schrieb Jarvis, Robert P. (Contingent): > > > > http://www.sunlabs.com/nova/cgi-bin/index.cgi?a=r&s=9343&e=935 > 6&f=%2Fhtdocs > >% 2Fself%2Flanguage.html&i= > > That link doesn't quite work, can you give me the real one? > > regards, > > nick > > - -- > "What's another word for Thesaurus?" > -- Steven Wright > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.2.1 (GNU/Linux) > > iD8DBQE+tlpf3P3nmXTVnC8RAjZ3AJwPtgZkuKOMfSb2v8URNuYdbZ9soQCfeLgh > 4+NbvhHPoYZeYAR24+vnIp4= > =Hu8W > -----END PGP SIGNATURE----- > > ********************************************************************** This message and any attachments are intended for the individual or entity named above. If you are not the intended recipient, please do not forward, copy, print, use or disclose this communication to others; also please notify the sender by replying to this message, and then delete it from your system. The Timken Company **********************************************************************

1 0

Re: *ListMorphs
by Daniel Vainsencher 05 May '03

05 May '03

It sounds right. I'm just saying you should remember to keep the model/morphic borders more or less where they are now. If you keep whisker's use of morphs, it is probably right. What I meant in detail - A domain object is the object that represents whatever your application is about. For example, in the browsers, the domain objects are lists of names of categories, methods, classes. Why lists of the names, and not of the class objects or method references? because PLM expects to get strings, it can't deal with most kinds of objects(*). So for a concrete example, when the users chooses to remove an object from a list, the domain object should be modified directly, and the UI should be updated from that. Is that clearer? (*) This is another limitation that it would be nice to be rid of. That was the idea I'd implemented - PLM's could be configured with an transformer object that can be sent a domain object, and produces an appropriate morph. This fixes both the "recieves only strings" and the "displays only string morphs" problems. Daniel tbr(a)mannynkapy.net (Tom Rushworth) wrote: > On Mon, May 05, 2003 at 01:48:53AM +0200, Daniel Vainsencher wrote: > > One thing you might consider is that your list should probably be of > > domain objects, represented by morphs, not a list of morphs. You might > > find the SM package ColorDevTools (CDT) intersting. > > I'll take a look, but you can probably help me reduce my ignorance a > little faster by elaborating on what you mean by "domain objects, > represented by morphs". I stole shamelessly from the Whisker Browser, > so the morphs in my list of morphs are relatively small wrappers around > the non-morph objects containing the information I want to display. Is > that what you mean, or am I hopelessly lost? > > > > I think this is generally an important direction, because more > > expressive lists would make many interfaces more useful. > > > > Daniel > > > > tbr(a)mannynkapy.net (Tom Rushworth) wrote: > > > On Sun, May 04, 2003 at 01:58:20PM +0100, Paul Chapman wrote: > > > > Are there any plans to build versions of the various list morphs > > > > which draw their rows directly instead of using a submorph for > > > > every row? For large lists the current implementation is very slow! > > > > > > There are some LargeList enhancements floating around. Check the > > > list archives and the swiki. My apologies for not having the > > > references handy. > > > > > > > > Cheers, Paul > > > > > > > While we're on the subject of ListMorphs, all of the list morphs seem > > > to expect the submorphs to be string/text morphs, and seem to provide > > > the submorphs with (effectively) infinite width. I want to display a > > > list of more complicated morphs where each submorph is a window with > > > 3 (or more) panes. I want the submorphs to be laid out proportionally > > > in the horizontal direction, but to be fixed size and scrolled in the > > > vertical direction. (For anyone who has used Quicken, think of how the > > > list of transactions in an account is displayed.) > > > > > > I just got such a morph working yesterday, but I had to make changes > > > all over the place :(. Now that I've proved to myself I can do it :), > > > I can ask the list: is there an obvious, easy way to do this? One that > > > doesn't involve a dozen new classes and changes to many old classes? > > > > > > Also, what should this morph be called? I've called it the > > > "PluggableMorphListMorh" in the sense of pluggable(MorphList)morph, not > > > because I like that name (I don't) but because I couldn't think of anything > > > better. Suggestions anyone? > > > > > > If anyone's interested, I'll post it once I get rid all the cruft and > > > dead end code from the things I tried that didn't work. Right at the > > > moment I don't think my pride could handle someone else looking at it :). > > > > > > Regards, > > > > > > -- > > > Tom Rushworth > > > > -- > Tom Rushworth

2 1

← Newer
1
...
62
63
64
65
66
67
68
...
71
Older →

Jump to page: