Squeak-dev August 2003

squeak-dev@lists.squeakfoundation.org

171 participants
757 discussions

Re: Extending FileList with CrLf
by Daniel Vainsencher 02 Aug '03

02 Aug '03

Jimmie Houchin <jhouchin(a)texoma.net> wrote: > Just for other outside information. Apparently the Python group has been > struggling with this issue. I normally don't read about Python, but > occasionally take a peek. :) > > This PEP has been implemented and is in the latest Python. > > http://www.python.org/dev/doc/devel/whatsnew/node7.html Note that they implemented the library so it requires an explicit parameter for it to enter "universal" mode. That is exactly what I'm proposing we do. This is not what CLFS does. > Talks about the problem and how they decided to handle it for Python. > > Personally, I'm for handling it as elegantly and transparently as > possible. I think that is what most people will expect. I understand > Daniel's position. However, Squeak is a high-level environment. I don't know what a "high level environment" is. Squeak applications are able to provide convinient application level abstractions because they build on rich lower level abstractions in the class library. I am stating that the two abstractions "line ending" and "auto detect line ending" should remain separate, so that different applications can each use what it requires. > Low-level options are there, but I believe most people expect intuitive > responses. I don't expect to open a text document written on the > platform I'm using Squeak on and Squeak not handle it properly. We could change the default to be OS line endings - this is a reasonable change. If you want your text editor to handle any strange line endings, change your text editor to use an "autodetect line endings" mechanism. However, the default should remain a clean "line ending" abstraction, without autodetection as it is in Python, C and probably most other general purpose languages. Daniel

5 5

IntegerPokerPlugin?
by Tim Rowledge 01 Aug '03

01 Aug '03

OK, what is this for, why should it be kept alive and who has any use for it? Currently it is _not_ in the VMMaker package and unless we hear a good reason for keeping its life support turned on it will be allowed to die with some dignity intact. tim -- Tim Rowledge, tim(a)sumeru.stanford.edu, http://sumeru.stanford.edu/tim Useful random insult:- Watching programs not listed in TV Guide.

3 2

Re: [ENH] SUnit-expected-failures-jf
by Daniel Vainsencher 01 Aug '03

01 Aug '03

I mean a way of naming and storing a bunch of related tests. TestSuite gives you the test composition abstraction, you need to store a TestSuite for the "expected failures", and then separately run the "expected failure" and the "regression" (everything else) parts of the test suite. This would require a little supporting UI and process to use, but it doesn't seem to require a lot of work. I don't know but I've been told that TestBrowser allows one to partition the tests more logically. Maybe it supports TestSuites somehow. Daniel Julian Fitzell <julian(a)beta4.com> wrote: > Can you expand more on what you mean by organizing structure? > > Daniel Vainsencher wrote: > > I think the need you're addressing could be better met with some > > organizing structure on top of TestCases... and that would be useful. I > > remember the mess it was to find a few score of RB TestCases not > > working, and then finding out that they are actually in implicit layers > > - the parser should be fixed before the refactorings, and so forth. > > Would have been nice if the layers were explicit. > > > > Daniel

3 2

RE: Startup time, a short review & discussion
by Dean_Swan＠Mitel.COM 01 Aug '03

01 Aug '03

Hi Andreas & John, For what it's worth, I have two Windows NT 4 machines and a Windows ME machine, and here are some very approximate (i.e. measure with a clock on the wall) startup times for Squeak3.2-4956 using Squeak 3.2.3/Tea 1.8 VM (release from Sep 5 2002, compiler gcc 2.95.2 19991024 (release). My idea of startup time is time from double-click on Squeak.exe until the Squeak window appears. CPU OS Physical Startup Time Bytecodes Sends RAM 1.4 GHz Pentium 4 WinNT4SP5 512M ~16 seconds ~119M/Sec ~2.90M/Sec 450 MHz Pentium II WinNT4SP6 256M ~18 seconds ~ 44M/Sec ~1.26M/Sec 700 MHz Pentium III WinMe 128M ~ 2 seconds ~ 67M/Sec ~1.94M/Sec These are very coarse measurements, but the two NT machines take about the same time to start up and the ME machine is much, much faster, while tinyBenchmarks more closely reflects the clock speeds of the CPU. Any ideas to explain this? -Dean "Andreas Raab" <andreas.raab(a)gmx.de> Sent by: squeak-dev-bounces(a)lists.squeakfoundation.org 07/31/03 08:00 PM Please respond to The general-purpose Squeak developers list To: "'The general-purpose Squeak developers list'" <squeak-dev(a)lists.squeakfoundation.org> cc: Subject: RE: Startup time, a short review & discussion > I've been glancing at squeak startup time on the mac this > morning and have a few observations. Actually I have noticed on > windows the images "SNAP' open, but take seconds on the mac and I'd > like to fix that... Another thought: Is it possible that you're allocating the memory lazily and the image loading process hits the exception handler upon first access to each memory page? This might slow down things quite a bit. In the Windows VM I am actually pre-allocating the memory needed for the image data (plus some extra) before the read comes along to avoid this. Cheers, - Andreas

3 2

Re: Help for Christopher Alexander in S.E. England
by David A. Smith 01 Aug '03

01 Aug '03

The term architect, as applied to computer science, was coined by Fred Brooks. He was referring to hardware architecture, but it quickly caught on to mean the design of any reasonably complex system in the field. DAS At 08:52 PM 7/31/2003 -0700, you wrote: >When computer people talk about design patterns of software engineering, >they are using a term coined by this Christopher Alexander. The term >architect, as applied to our field, was similarly coined by him. He >appeared at OOPSLA as a guest speaker. In short, the man has exciting ideas. > >At 11:31 AM 7/31/03, you wrote: > >>Christopher Alexander, author of 'A Pattern Language', wants to write a >>simulation to demonstrate the ideas in his new series of books 'The Nature >>of Order' (the first volume is out, and truly outstanding). I have >>persuaded him that Squeak is the right language to do it in and he's hooked. >>Chris has not programmed for some years (though he is a mathematician, >>originally) and needs some tutorial help on the basics of the language. >> >>He'd like to find a fluent Squeaker who'd give him a few hours of voluntary >>time. Chris is currently living on the South coast of the UK (West Sussex). >>The ideal would be a grad student in Southampton or Portsmouth Universities. >>(Anyone know of any Squeakers there I could contact directly?) If this >>doesn't work, someone on the end of a phone would be acceptable. >> >>This is an opportunity to meet and interact with one of the world's greatest >>contemporary thinkers. I'm sure we can help him out. (I've offered my help >>on the OO design side, but I'm not a sufficiently fluent programmer to give >>him the basics he needs at the beginning.) Just to make it clear, he wants >>to write the simulation himself, this is not an attempt to get someone to >>write it for him. >> >>If you can help, or know any contacts who might, please get back to me >>directly and I'll put you in touch with him. >> >>Richard Pawson >

2 1

[FIX] SystemOrganizer-remove-empty-category-fix-jf
by julian＠beta4.com 01 Aug '03

01 Aug '03

Here we go a second time with a fix that actually works. from preamble: "Change Set: SystemOrganizer-remove-empty-category-fix-jf Date: 1 August 2003 Author: Julian Fitzell When you remove a category in in the Browser, SystemOrganizer removes all empty categories for you. Surely this isn't the correct behaviour? Can't believe I've never noticed this before..."!

1 0

Re: [ENH] SUnit-expected-failures-jf
by Daniel Vainsencher 01 Aug '03

01 Aug '03

I might find it interesting. Please keep us posted. Daniel Julian Fitzell <julian(a)beta4.com> wrote: > Hmm... we're using TestBrowser and it doesn't seem to do anything > special about TestSuites - just lets you run categories of tests easily. > > I suppose you could set up a TestSuite for failing tests, but unless you > actually remove the tests from their proper TestCases, running that > TestCase will still cause errors. Then I suppose you can create a > TestSuite for passing tests, but you have still lost the flexibility of > being able to run individual TestCases. And I don't like the idea of > actually removing the test from the TestCase because you don't know > which TestCase it needs to go back to when it passes again. > > It feels like this is another case where we need meta-information > associated with methods - but the best workaround I know of is to have > another method hold the meta-information. > > Still - better tools would make a big difference however we do it - when > I have time to look at this further, I would definitely look at the > testing UIs. Is there any interest in having a solution to this problem > or will I just keep it in the realm of Personal Hacks? :) > > Julian > > Daniel Vainsencher wrote: > > I mean a way of naming and storing a bunch of related tests. TestSuite > > gives you the test composition abstraction, you need to store a > > TestSuite for the "expected failures", and then separately run the > > "expected failure" and the "regression" (everything else) parts of the > > test suite. > > > > This would require a little supporting UI and process to use, but it > > doesn't seem to require a lot of work. I don't know but I've been told > > that TestBrowser allows one to partition the tests more logically. Maybe > > it supports TestSuites somehow. > > > > Daniel > > > > Julian Fitzell <julian(a)beta4.com> wrote: > > > >>Can you expand more on what you mean by organizing structure? > >> > >>Daniel Vainsencher wrote: > >> > >>>I think the need you're addressing could be better met with some > >>>organizing structure on top of TestCases... and that would be useful. I > >>>remember the mess it was to find a few score of RB TestCases not > >>>working, and then finding out that they are actually in implicit layers > >>>- the parser should be fixed before the refactorings, and so forth. > >>>Would have been nice if the layers were explicit. > >>> > >>>Daniel > > > >

1 0

[ANN] New Free Online Book: The Art and Science of Smalltalk
by Stephane Ducasse 01 Aug '03

01 Aug '03

Please distribute the following information: Hi, I'm happy to let you know that we got a new book online: The Art and Science of Smalltalk by Simon Lewis Thanks simon and Prentice-Hall. The scanning was done by A. Lienhart and I spent the night ocring it. Enjoy it (some pages are missing as indicated in the file name we are fixing that). The cache of my web page does not allow me to see if you can see it. http://www.iam.unibe.ch/~ducasse/WebPages/FreeBooks/Art/ I have now little Smalltalk to OCR, and the rights for Inside Smalltalk and the Smalltalk-80 books :)....soon available. Stay connected. Stef, so called a troll in the past. 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 1

[FIX] SystemOrganizer-remove-empty-category-fix-jf
by julian＠beta4.com 01 Aug '03

01 Aug '03

from preamble: "Change Set: SystemOrganizer-remove-empty-category-fix-jf Date: 1 August 2003 Author: Julian Fitzell When you remove a category in in the Browser, SystemOrganizer removes all empty categories for you. Surely this isn't the correct behaviour? Can't believe I've never noticed this before..."!

2 1

Re: [ENH] RemovePackageMethod
by Daniel Vainsencher 01 Aug '03

01 Aug '03

My timing shows 2 minutes to load, and .5 a minute to load Morphic (which is quite extraordinarily large, because it includes both the framework(s?) and the applications). But the question in my mind is not just whether these big chunks of image can be merged quickly, but how this can become a basis for developing them. I have more questions than answers on that count... Daniel Avi Bryant <avi(a)beta4.com> wrote: > > On Fri, 1 Aug 2003, Bernhard Pieber wrote: > > > Ah, right! I forgot that there is code which is not in a package. How > > inconvenient! ;) > > > > Hmm, wouldn't it be easy to convert the existing system categories to > > Monticello packages. Then you could just reload any of them in case you > > needed. I have to think about that. Probably I am again forgetting > > something very basic. It's been a long day. > > Yes, I think it would be a good idea to start experimenting with that. > Managing huge packages like Morphic may require some profiling of the > Monticello code; I know Daniel did some speed tests and found that a merge > of a package of that size currently takes a few minutes. > However, if we can get that running a bit better, I don't see why we > couldn't put up a repository with Kernel.mcv, Collections.mcv, > Network.mcv, and so on. > > Avi

1 0

Jump to page: