Squeak-dev January 2002

squeak-dev@lists.squeakfoundation.org

176 participants
612 discussions

Re: Project Publishing issues
by Bob Arning 03 Jan '02

03 Jan '02

Hi Ross, On Wed, 02 Jan 2002 19:20:34 -0800 Ross Boylan <RossBoylan(a)stanfordalumni.org> wrote: >I've noticed the following problems when I publish a project: > >1. Several objects in the project are related through the #when:send:to >event mechanism. Not all the dependencies survive intact, though it looks >as if some do. The one that didn't involved one of the published Morphs as >a receiver; the others had non-graphical receivers. It may also matter >that the source of the notifications kept its dependencies in the >system-wide EventsFields, rather than its own instance variable (like >EventModel). >From a quick look at the code, events registered on EventModel or Morph subinstances should be exported correctly, but events registered on other kinds of objects likely will not be. Similarly, dependents registered on Model subinstances should be handled correctly while others would not be. I would suggest checking to be sure which events/dependencies you are losing and perhaps implementing #myEvents and/or #myDependents for those classes in a manner like the current non-Object implementations. >2. The key data are rooted in a class variable of one of the project's >classes. The data made it across, but the variable is nil. A class variable is a handy place to stash things that have no other logical home, but they will be missed in the project exporting mechanism. Why not store the value as a property of the world in which your project lives? That way it will be exported and imported without problem. >Can anyone suggest a work around or fix for these problems? I think I >remember someone else commenting that publishing can break >dependencies. I'm not sure how it works for update/changed stuff. > >Is there a way to get prolog and cleanup code into the project (e.g., in >its change set)? If so, I could try (for problem 2) something horrible like >MyClass setClassVariable: MyClass allInstances first >since there's only one instance. I'm not sure if things would be in a >proper state for it to work, though. You can put a postscript on the project change set, but this will be run before the project itself has been loaded, so there will not be any instances to refer to. You could change the method that accesses this variable from getClassVarRoss ^ClassVarRoss to getClassVarRoss ^ActiveWorld valueOfProperty: #ClassVarRoss ifAbsent: [ ActiveWorld setProperty: #ClassVarRoss toValue: ClassVarRoss. ClassVarRoss ]. This way, the varible will be obttained from the world, except for the very first time when it will come from the class variable slot. > > >3. I also found an interesting problem and solution. I had a class accessor >MyClass>>ross > ^RossInstance >where RossInstance was a global variable. When I tried to read the >resultant project in I got an unrecognized key error. I got around this by >deleting the method from the change set. (It seems likely that rewriting >to Smalltalk at: #RossInstance might fix the problem also.) I think the moral of several of these issues is to try to avoid globals. It's probably better for objects that need to refer to another object to have an instance variable for it or a way to talk to another object that does. Using Smalltalk itself or classes to serve this purpose does not mesh well with the project exporting mechanism. > >4. Is there a way to move a project you have published to your local disk >to the server without going through the entire publishing process? I'm >publishing locally to test that it's working. When it is, it would be nice >to be able to ship the project file off directly. That depends on the server, but it's probably not worth the trouble. The code you would need to emulate is toward the end of Project>>exportSegmentWithChangeSet:fileName:directory: and there are a number of steps required to get it all lined up correctly. Cheers, Bob

1 0

mmap versus read in MacOS X
by Eric Scharff 03 Jan '02

03 Jan '02

The read time for loading the image into memory seemed a little high to me, so I wrote a tiny benchmark. The simple program attached can be used to compare reading performance of block reads using read(2) versus mmap(2). It counts the number of A's in the file specified on the command line. On my G3/400 running MacOS 10.1.2, read and mmap performance seem roughly the same. On a 75MB file, this took between 2.2 and 2.7 seconds to run. (Note that my tests include rebooting to avoid cache hits.) The interesting thing is that the although the read solution takes spends far more time making system calls (1200ms versus 260ms), the clock time is roughly the same. Note too, that on my G3/400, reading a 75MB file takes only 1280ms. How does this compare to Squeak's image reading time? -Eric

2 1

Re: Type Inferencer to the Rescue
by Lex Spoon 03 Jan '02

03 Jan '02

> > So I would like to know what is the status of the type inferencers? > Is Lucid still developed? I'm still chugging along on Lucid. It does find types for a good many expressions, but I'm trying to improve on it. I haven't reposted a newer version in a while -- first, the next version will be better, and second, it's probably not real fun to play with right now, because of the way it deals (or doesn't deal) with code changes. Specifically, on that last note, I've started assuming that the following things are available to the analysis: 1. parse trees for all methods 2. a senders-of table, to avoid searching through the system 3. an implementors-of table, again to avoid searching What I haven't done, is arranged for these tables to get updated whenever someone changes code in a browser. It's more work than it seems at first, because there are several different ways someone can change code (add a method; rename a class; reparent a class; etc.). And there's a temptation to generalize, so that other tools could react to changes in the system code.... -Lex

3 3

[ANN] swiki page on Cleaning up junk
by Ross Boylan 03 Jan '02

03 Jan '02

I've collected what I've learned about creating, identifying, and cleaning up cruft in an image in a page on the swiki at http://minnow.cc.gatech.edu/squeak/2176 The link to it is from the "Practical Wizardry" page. I hope it is useful. Please let me know of any comments or errors.

1 0

UIUC server troubles?
by Ross Boylan 03 Jan '02

03 Jan '02

This morning I got a time out trying to go to ftp://st.cs.uiuc.edu/pub/Smalltalk/Squeak/, the address given by http://minnow.cc.gatech.edu/squeak/1582. I tried later with same result. Just now, I finally succeeded. However, trying to update a 3.1ish image, I get the error no new updates on squeak.cs.uiuc.edu/Squeak2.0/. I don't know why it's looking at 2.0, and I know there were some changes to the updates servers, but ..... is anything wrong with the server?

2 1

Re: [ANN][VM][SF] my Unix patches integrated into SF layout
by Lex Spoon 03 Jan '02

03 Jan '02

> > I noticed that sq.h is taken from the build dir hierarchy, by the way, > so it's not entirely consistent... I think that's the proper location to grab this file from. The sq.h in the Cross-platform directory probably shouldn't even be there, really. My understanding is: src - holds stuff generated by Squeak Cross - holds stuff that all platofrms can use, that wasn't automatically generated unix - holds Unix-specific stuff Actually, though, src ends up combining all three, the way VMMaker works. This combining behavior seems to be a little tricky to get just right, however.... > Or at least I don't understand what is pulled from where, exactly. > And I get warnings about duplicate rules in the Makefile. Could you try deleting platfoms/Cross/SurfacePlugin/SurfacePlugin.c, running configure again, and seeing if the errors go away? > I'm not courageous enough to dive into this autoconf/Makefile stuff to > find out why or do something against it, though... I hear you! autoconf stuff is a mess even at its best! > However, I have attached a new version of sqUnixMemory.c, which removes > a few pointer/int cast compiler warnings. Briefly tested with/without > mmap enabled. Awesome, those are annoying! I've uploaded your version now, along with some small changes to sqPlatformSpecific.h I'm sure you made. -Lex

2 1

Project Publishing issues
by Ross Boylan 03 Jan '02

03 Jan '02

I've noticed the following problems when I publish a project: 1. Several objects in the project are related through the #when:send:to event mechanism. Not all the dependencies survive intact, though it looks as if some do. The one that didn't involved one of the published Morphs as a receiver; the others had non-graphical receivers. It may also matter that the source of the notifications kept its dependencies in the system-wide EventsFields, rather than its own instance variable (like EventModel). 2. The key data are rooted in a class variable of one of the project's classes. The data made it across, but the variable is nil. Can anyone suggest a work around or fix for these problems? I think I remember someone else commenting that publishing can break dependencies. I'm not sure how it works for update/changed stuff. Is there a way to get prolog and cleanup code into the project (e.g., in its change set)? If so, I could try (for problem 2) something horrible like MyClass setClassVariable: MyClass allInstances first since there's only one instance. I'm not sure if things would be in a proper state for it to work, though. 3. I also found an interesting problem and solution. I had a class accessor MyClass>>ross ^RossInstance where RossInstance was a global variable. When I tried to read the resultant project in I got an unrecognized key error. I got around this by deleting the method from the change set. (It seems likely that rewriting to Smalltalk at: #RossInstance might fix the problem also.) 4. Is there a way to move a project you have published to your local disk to the server without going through the entire publishing process? I'm publishing locally to test that it's working. When it is, it would be nice to be able to ship the project file off directly.

1 0

Re: Type Inferencer to the Rescue
by Lex Spoon 03 Jan '02

03 Jan '02

> > 1. parse trees for all methods > > In Oasis I've only needed these once per method- after having done some model > building, I can discard these. My experience was that these took up a HUGE > amount of memory, so dropping them was nice. The model I build that contains > such information ( see below ) is much much smaller than the parse trees. Yes, they do take up tens of megabytes. Very sad. However, for many reasons, I'm doing it anyway. Here's a few: 1. It needs the parse trees while it's running, anyway, and so you still need the RAM. 2. It can be optimized later. 3. Type inference is well worth 50 MB of RAM. 4. Memory is cheap, anyway. 50MB is somewhere around the price of 10 gorditas. > > 2. a senders-of table, to avoid searching through the system > > 3. an implementors-of table, again to avoid searching > > I don't use tables, but this sort of functionality is an explicit part of what > could be called a message trace model. That model should be useable by lots > of different analysis tools, not just the ones in Oasis. Awesome. I can't wait! -Lex

1 0

Re: VMMaker thoughts
by Lex Spoon 03 Jan '02

03 Jan '02

> In that case you must surely be running a very out of date version. The > latest stuff has had VMMakerTool in it for several months now. It is in > http://sumeru.stanford.edu/tim/pooters/SqFiles/deltas/VMMaker-3-1-version3.… Okay, I had 3-1-version 2.5. I'll have to download and check the newer version.... -Lex PS -- "months" ago?! several months?! heheh, open source sure flies sometimes!

1 0

Re: getting really weird: Player39 in PointerFinder
by Bob Arning 03 Jan '02

03 Jan '02

On Wed, 2 Jan 2002 01:57:17 -0800 Ross Boylan <RossBoylan(a)stanfordalumni.org> wrote: >On Tue, Jan 01, 2002 at 06:18:27PM -0500, Bob Arning wrote: >> On Tue, 1 Jan 2002 14:19:45 -0800 Ross Boylan <RossBoylan(a)stanfordalumni.org> wrote: >> >I don't know how Etoy scripting ended up with one of my windows as an owner... >> >> Ross, >> >> Is there a method #script1 in Player39? What does it say? This may give a clue as to how things got started. >> >> Cheers, >> Bob >> >> > >Yes. In full: > >script1 > self setHeading: -4 > >There is a similarly exciting script2. Is this a rotation command? Yes. >Possible scenario: while trying to understand some of the new stuff, I >created a tile for my window. Then I clicked a bunch of the controls >to see if the window would move around. > >If I did that, which is likely, would that create these Player classes >and the corresponding script owners? Yes. >How come closing the tile >doesn't destroy the class? Do you expect ordinary classes to be deleted when you simply close a browser? Off the top of my head, I don't think any automatic removal of these classes exist, but you can look at Player class>>freeUnreferencedSubclasses and Player class>>abandonUnnecessaryUniclasses for some ideas on how to get rid of them. Cheers, Bob

2 1

Jump to page: