Squeak-dev January 2008

squeak-dev@lists.squeakfoundation.org

147 participants
192 discussions

Unix VM file locking?
by Andreas Raab 28 Jan '08

28 Jan '08

Hi - Had a fun morning this morning with some customer trouble which was partly caused by the Unix VM not using write locks. E.g., the Unix VM will happily do things like: f1 := FileStream newFileNamed: 'foo.bar'. f1 nextPutAll: 'Hello World'. f2 := FileStream oldFileNamed: 'foo.bar'. f2 nextPutAll: 'Oh, noes'. etc. Is there an easy way to patch the Unix file code to avoid this, e.g., behave like Windows in such that it would disallow opening the same file for multiple writes? Cheers, - Andreas

6 7

YAXO latest broken?
by Keith Hodges 28 Jan '08

28 Jan '08

To whom it may concern, When using the latest XML-Support/YAXO with universes it is sending the image into an un-interruptible state when it tries to parse the xml returned form the server. cheers Keith

2 1

Re: Real profiling: Why is my UI so slow?
by Jerome Peace 28 Jan '08

28 Jan '08

Real profiling: Why is my UI so slow? Hi Matthew, I have seen some interesting problems with the image and browser speed. The slowness seems to come when the browser is updating itself and that can be triggered by other morphs overlapping (e.g. other browsers). See the mantis report: http://bugs.squeak.org/view.php?id=2084 0002084: Each overlaping browser slows down screen updates considerably. Caching the image would speed things up. This is mostly external observations. At the time I reported this I hadn't learned to use the tally tools usefully. (I still haven't. They give too much information. I want something that will point me quickly to the culprit.) My observation is that caching images would be one way to speed things up. I haven't tried to tackle the problem directly. (Except to make 3.9 workspaces start out smaller. The less things overlap the smaller the problem.) Animated morphs are a good way to aggrivate the problem. By stepping and invalidating their bounds. Hmm, I wonder if a one pixel rectangle would be good for pinpointing the worst update areas?.Invalidate the rectangle each step. Move it around the screen to see which pixel causes the greatest slowdown. The step code could note the clock interval between steps. Color a map of the screen with the results. Alright, we have a crazy idea. Lets see where it gets us. I'll post something to mantis when it gets interesting. Cheers. Yours in curiosity and service, --Jerome Peace *** >Matthew Fulmer tapplek at gmail.com >Sat Jan 26 06:25:35 UTC 2008 > > >Hi all. While playing around with my DeltaStreamBrowser, I see >that it takes four seconds for the UI to update after I click on >a change. And I have no idea why. > >I've often wondered why the browser is so slow too, and now I >have the need to find out. So, How does one go about debugging >this? > >I know about MessageTally and TimeProfileBrowser, but I've only >seen them used on well-defined code, not in the wild in the >Squeak UI. And I know about world > debug > Message Tally, but I >really have no idea what the numbers they report mean, or what >to look for in the reports they return, or how to even know what >messages were sent while actually processing my click vs. in >between while I move the mouse around. > >What I would really like to happen is, I tell squeak: "Pop-up a >debugger on the next click". Then, on the next click, the UI >process is suspended and shown in a debugger that has it's own >private UI loop. Is anything like this possible? I am sure the >Morphic event loop was not created without the help of the >debugger, but I really have no idea how to trace a UI event from >post to screen-update. > >So, that is what I have to work with, and how I don't know what >to do with it. So, how do I track down the latency problem? *** ____________________________________________________________________________________ Looking for last minute shopping deals? Find them fast with Yahoo! Search. http://tools.search.yahoo.com/newsearch/category.php?category=shopping

5 4

Unix vm on Mac OSX 10.5 uses X11
by Keith Hodges 27 Jan '08

27 Jan '08

I just discovered on Mac OS X 10.5 Leopard that the unix vm (3.9-8/3.9-10) uses leopards X11 support rather than the quartz interface. When I start up using -quartz I get the following message. 2008-01-27 17:32:47.342 unixsqueak[6057:10b] Warning once: This application, or a library it uses, is using NSQuickDrawView, which has been deprecated. Apps should cease use of QuickDraw and move to Quartz. cheers Keith

1 0

who ever performed bit logic on large negative integer?
by nicolas cellier 27 Jan '08

27 Jan '08

SmallInteger < 0 are stored directly in two complement native format (well, except the tag bits) But LargeNegativeInteger are stored same as their absolute value. This does simplify some algorithms. But this leads to quite tricky ones when bit logic is to be performed based on two complement. And whenever tricks are used, bugs frequency generally tends to increase. I believe long life bugs are lying in such swamps. That's why i'm very proud of finding accidentaly http://bugs.squeak.org/view.php?id=6874 via http://bugs.squeak.org/view.php?id=6873. As usual, most of you won't ever care of it. Yeah, one more useless bugfix i am specialized in since the probability you bump into it must not exceed 1.0e-8 per hour of Smalltalking. But who knows, it's a kind of insurance in case a silly engineer had the idea to run nearest nuclear plant in Squeak! And well, it's also for the beauty of mathematics! Nicolas

5 10

[ANN] SqueakPeople Certifications
by Germán Arduino 27 Jan '08

27 Jan '08

Hi: Trying to test the newest versions of Squeak (3.10) and Seaside (2.8) I developed a tiny app to group the SqueakPeople by its certifications using the page of SqueakPeople with this data, as data source. The app is running on seasidehosting, at this link: http://squeakpeoplestats.seasidehosting.st/seaside/SqueakPeopleStats Enjoy :) -- Germán S. Arduino http://www.arsol.biz http://www.arsol.net http://germanarduino.blogspot.com

3 4

Vats, Islands, and Collections: Finding cheap parallelism
by Matthew Fulmer 27 Jan '08

27 Jan '08

I'm going to be working on making squeak multi-core as part of my grad work. I have a bit of work to do before I get to the stage where I actually have to implement a parallel code model, but I am thinking about it in the abstract as a mental exercise for now. I will most likely use the vats/islands system used by SqueakELib and Croquet, as it has been validated. But I have a pressing concern: How could this be used to do what seems (to me and my mentors) to be the cheapest, simplest, most obvious parallelism extraction ever: get Collection>>do:, collect:, etc. to run one element per native thread (= vat = island). However, I may be overlooking copying costs (code, objects, or both) that make this more a naive parallelization than I expect. Assuming the elements are read-only for the duration of the enumeration (the common case), they can be safely copied willy-nilly to whatever vat they need to be in to run (or use shared memory if it is available and fast). Can this be done using vats? Does copying kill it? I get asked about this problem every time I talk about vats. Everyone seems to think this (loop-level parallelism) is the simplest, most fool-proof way to make things faster on a multi-core system, and give me funny looks when I don't have a quick answer on how this works in the vats model. -- Matthew Fulmer -- http://mtfulmer.wordpress.com/ Help improve Squeak Documentation: http://wiki.squeak.org/squeak/808

12 15

Smalltalk UK meeting on Saturday 8th March
by bryce＠kampjes.demon.co.uk 27 Jan '08

27 Jan '08

We're organising a Smalltalk event in London. This is going to be on a Saturday afternoon, we'll probably head to a pub afterwards. We'll aim to get a room with a projector. Details are here: http://www.xpdeveloper.net/xpdwiki/Wiki.jsp?page=SmalltalkUK20080308 RSPV required as it's being hosted in corporate offices. Please add your name if you're interested in either attending or presenting. Bryce P.S. Sorry about sending this twice. This time should have the subject line.

1 0

squeak releases, now and in the future (reply to Matthew)
by Jerome Peace 27 Jan '08

27 Jan '08

squeak releases, now and in the future (reply to Matthew) Hi Matthew, I don't have a direct reply to your post. I do want to thank you for replying. The level of information was great. I need to think about it rather than respond. I do want to pick up on just one thing that you mentioned. Matthew Fulmer tapplek at gmail.com Mon Jan 21 22:41:15 UTC 2008 >Should the release team focus on one facet at the expense of the others? No! What I am searching for is a sense of what the role and purpose of the release team is and will be. Right now I am uncomfortable with the lack of communicated direction. (Emphasis on the word communicated.) This thread and your response have helped inform. I've also pick up some clues by looking at the logs of the irc channel. Thanks. Yours in service and curiosity, --Jerome Peace ____________________________________________________________________________________ Be a better friend, newshound, and know-it-all with Yahoo! Mobile. Try it now. http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ

1 0

squeak releases, now and in the future (reply to Keith)
by Jerome Peace 27 Jan '08

27 Jan '08

squeak releases, now and in the future (reply to Keith) *** > >Keith Hodges keith_hodges at yahoo.co.uk >Mon Jan 21 17:00:41 UTC 2008 > >Jerome Peace wrote: >> Hi all, >> >> I think we are at the anniversary of the start of the >> six month timebox for 3dot10. >> >> I am watching a lot of interesting developments. >> Damiens efforts seem to have become squeak's flagship. >> And VPRI has found resources and interest in >> developing >> squeaklands branch of 3.8 and OPLC. >> >> As for our squeak-org branch, it seems to me the basic >> maintenence and releases of squeak are in neglect. >> >I do not entirely agree with this conclusion since there is work taking >place in this arena. > >The LevelPlayingField initiative is aiming to provide a >platform/framework whereby we can care for all of the squeak.org >releases that we use. This also provides one framework for integrating >portions of work towards future releases. > >Within LevelPlayingField there are spaces for several projects that will >contribute to a future release, and so really there is no stagnation at >all, and there is plenty of room for many others to contribute. Hmm, I was being vague and hinting at a problem without specifing it exactly. You are correct that your work is addressing some of the problems that concern making new releases. Perhaps I need to claim a distinction between work on tools and the organizational task of focusing resources on the process of creating (and blessing) releases. Its the second task that is not happening. And it is bothering me that its not being talked about (goran excepted) or acknowledged. At least not in the mailing list. >We have: > >1. DeltaStreams work by Matthew, including debugging of the SystemEditor >for atomic loading. >2. Monticello can also use SystemEditor (traits not yet supported) >3. http://installer.pbwiki.com/MinorFixesUnstable >4. http://installer.pbwiki.com/Clean > >There are other tool based projects on the go as well, but they are not >yet ready to discuss publically. >> My druthers are that time boxes get honored. That >> >I myself am not such a stickler for timeboxes, I think you could >effectively time box bug-fix releases. Timeboxes are important because they provide both focus and a way to limit scope. Untimeboxed, efforts tend to increase scope towards the end of the project. Projects generate learning. Learning generates new ideas. New ideas generate new scope. New scope at the end of a project generates mischief. Andreas suggested that the release process in any one cycle only consider fixes and changes that are already to go at the beginning of the cycle. I believe that would give a release team a good way to judge what they could do in a six month period. Keeping to the timebox and having a future release team in the wings for the next six month period would insure that ideas generated during that time are addressed. Right now this seems desirable to me but not achievable. >I do believe that improvements in >tools will make everything a lot easier. The tools that need improvement >are not part of the squeak kernel/core so it is not surprising that the >kernel is not being looked after as perhaps it might be until those >tools are complete. Yes, I agree. For updating a image MC is an out of sequence tool*. It was not meant for such work and release teams should not rely on it as much as they have chosen to do. >> future release teams be sought. And that the current >> release be wrapped up and delivered, lessons learned >> and new development planned for. >> >> Where do we go from here?* >> >I suggest that anyone who is interested in contributing to 3.11 >contribute to and join the #squeak irc channel to join in with others >who are like minded and who are working on the future of squeak in >several different directions. (Matthew, Goran, Craig, Gulik etc etc) Touché. I have not used the irc channel. I'll heed your suggestion an look into it. >> What endevors are worthwhile? >> >We can put together ideas and use LevelPlayingField as a place to slot >together sub-projects. >> What resources are available for future efforts? >> >I then propose that the "official" release team be comprised of people >who have been taking an active role in the process, and who use #squeak >irc communications regularly so as to encourage many more contributors >and to faclitate online teamwork. Gjallar has used this model reasonably >successfully. > Cool. > >best regards > >Keith > >p.s. For anyone itching to contribute immediately, the "Clean" script >currently leaves 12 obsolete classes in the system, so these need >tracking down ;-) > *** Yours in curiosity and service, --Jerome Peace *out of sequence tools cause there to be a lot of backtracking. Even small errors in 3.10 required extreme efforts to back off of because of the difficulties in using MC. Large portions of the effort consisted of Edgar educating himself as to the various pitfalls and remedies. The focus of the release team should be on what needs to go into the image (or be taken out). Not on how to do it. For more on the importance of sequence check out Christofer Alexanders writings. Amazon reviewer: "... Book 2 in this series goes in depth into the concept that things can only be built to enhance life and be truly beautiful and useful if they are built in a sequence of appropriate steps. ...." Useful link: http://www.patternlanguage.com/ ____________________________________________________________________________________ Looking for last minute shopping deals? Find them fast with Yahoo! Search. http://tools.search.yahoo.com/newsearch/category.php?category=shopping

1 0

← Newer
1
2
3
4
5
6
7
8
9
...
20
Older →

Jump to page: