Squeak-dev September 2005

squeak-dev@lists.squeakfoundation.org

157 participants
266 discussions

OmniBase + Linux
by Ramiro Diaz Trepat 22 Sep '05

22 Sep '05

Hello everyone, I've tryied OmniBase on Linux without success. When I run the tests in OmniBaseTester I always get an exception relative to locking: ODBLockxxxx. This does not happen in Windows. Is there a particular version of the VM or image to run OmniBase on Linux ? Thanks in advance. -- http://neosmt.com

2 1

Compiling the VM statically
by Lukas Renggli 21 Sep '05

21 Sep '05

Hi, I am trying to compile the Squeak 3.7 VM on Unix statically, so that it doesn't require any external libraries: all plugins and external libraries (libc, libm, ...) should be linked into the resulting binary itself. I tried: export CFLAGS=-static export LDFLAGS=-static ../platforms/unix/config/configure --enable-static make However I get: bash:~//Squeak-3.7-7/build# ldd squeak libutil.so.1 => /lib/tls/libutil.so.1 (0xb7fdf000) libdl.so.2 => /lib/tls/libdl.so.2 (0xb7fdb000) libm.so.6 => /lib/tls/libm.so.6 (0xb7fb9000) libnsl.so.1 => /lib/tls/libnsl.so.1 (0xb7fa5000) libc.so.6 => /lib/tls/libc.so.6 (0xb7e70000) /lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0xb7fea000) What am I missing? Thanks for any hints, Lukas -- Lukas Renggli http://www.lukas-renggli.ch

2 1

How much is my Squeak Image like Sourdough bread?
by Austin King 21 Sep '05

21 Sep '05

Greetings Squeakers, I am interested in the history of the Squeak 3.8 image. Metaphorically how much is my Squeak image like sourdough bread? Literally, are there instances of objects that have been living ( or allocated ) in the image for 30 years? I can see how the individual class and method definitions can be traced back, just like libraries in other file based programming languages, but the runtime system of the image seems wholly different. Was the image thrown away during major transitions in VM such as Smalltalk-72 -> Smalltalk-80 -> Squeak? If not, how is this transition made? Thanks, Austin P.S. I'm coming from a J2EE perspective, where these cultural facts are quite enlightening. Deploying a content change recently required a server reboot.

3 2

REMINDER: Squeak Chat Thursday September 22nd 17:00 GMT
by Ken Causey 21 Sep '05

21 Sep '05

Just a reminder that the next Squeak Chat will take place between 17:00 (5 PM) and 21:00 (9 PM) GMT Thursday September 22nd, 2005. As usual we will meet in #squeak on irc.freenode.net. More information about the #squeak IRC channel can be found at http://people.squeakfoundation.org/article/7.html This is a good time to contribute to the development of Squeak by participating in the Harvesting Process. One of the main ways to contribute is by Reviewing submitted fixes and enhancements. For information about the Harvesting Process see http://minnow.cc.gatech.edu/squeak/3152 If you have the time please attend and work toward the future of Squeak. The next seven chats are scheduled for 21:00 (09:00 PM) GMT Monday September 26, 2005 01:00 (01:00 AM) GMT Saturday October 01, 2005 05:00 (05:00 AM) GMT Wednesday October 05, 2005 09:00 (09:00 AM) GMT Sunday October 09, 2005 13:00 (01:00 PM) GMT Thursday October 13, 2005 17:00 (05:00 PM) GMT Monday October 17, 2005 21:00 (09:00 PM) GMT Friday October 21, 2005

1 0

Third Way...
by Mayuresh Kathe 21 Sep '05

21 Sep '05

Hi, Is the author or Third Way still around on this list? ~Mayuresh

17 41

[ANN] Smalltalk job offer in France
by Noury Bouraqadi 21 Sep '05

21 Sep '05

Hello, Please find below an offer for both junior and experienced Smalltalkers in France. Regards, Noury ------------ INGENIEUR D’ETUDES SMALLTALK Nous sommes une SSII à taille humaine intervenant dans le domaine des systèmes d’information partenaire de grands comptes banque / assurance en maîtrise d'oeuvre et en assistance à maîtrise d'ouvrage. Nous recherchons pour un compte tertiaire un analyste SMALLTALK afin de participer à un important projet de refonte d’application. Vous aurez en charge, au sein d'une équipe, les phases de conception, réalisation et tests. De formation Bac+4/5, doté d’un bon relationnel, vous avez de bonnes capacités d’adaptation ainsi qu’un bon esprit d’analyse. Deux profiles nous intéressent : -jeune diplômé : ayant effectué un stage significatif (environ 6 mois) en Smalltalk -expérimenté : vous justifiez d’une expérience d’au moins 2 ans en conception objet et développement SMALLTALK. Contact Karin JENCK Chargée de Recrutement GTI Consultants Tel : 01 53 70 71 34 E-mail : kjenck(a)gti-info.com -------------------------------------------------------------- Dr. Noury Bouraqadi - Enseignant/Chercheur Ecole des Mines de Douai - Dept. G.I.P http://csl.ensm-douai.fr/noury European Smalltalk Users Group Board http://www.esug.org Squeak: an Open Source Smalltalk http://www.squeak.org --------------------------------------------------------------

1 0

eCompletion, 0001678 and keymapping
by Stéphane Ducasse 21 Sep '05

21 Sep '05

Hi guys we (the hauswarterin = caretaker = lonely harvesters) would like to have in 3.9 as external packages but certain loaded together: eCompletion keymapping and the enh of romain for navigating menus So it would be nice if you guys could have a look at the latest version of 3.9alpha (upgrade the packages) Monday we will produce a new release and check how your stuff could work nicely together and identify what should be push in the kernel to make your package life easier. Stef

4 4

OT: sed script help, please
by Tim Rowledge 21 Sep '05

21 Sep '05

I need a sed script (to run on OS/X Tiger) that will take an exported email file and convert it to a variant of mbox format that OS/X Mail will import. I didn't know until I tried just how many variations and nit-picky hassles there could be for this apparently simple action - export email in 'From' format (ie mbox) and import to Mail. Sheesh. What fun. So far as I can discern I need to export as rmail format in order to get an easily searchable message delimiter; the mbox esport seems to not get it quite right. Did I say 'what fun' already? Once I have an rmail file it seems to work if I convert the #!rmail 0000000 lines to >From foo {some date sort of stuff} since everything after the From is ignored except that it has to be there or the whole message or even file is ignored! OS/X has sed sitting there and that seems like a sensible tool for converting several umpty-bytes of email rather than trying to use my text editor or even some Squeak script. Except that I don't know sed and I'll bet somebody out there does and could solve this in a small number of seconds. Help most gratefully received. tim -- Tim Rowledge, tim(a)rowledge.org, http://www.rowledge.org/tim Useful random insult:- The perfect personality to write software manuals.

6 5

Re: Image using high CPU in WeakIdentityKeyDictionary>>scanFor:
by Chris Muller 21 Sep '05

21 Sep '05

> >>A TimeProfile while the system was doing nothing showed virtually all > >>(99.4%) the time going to WeakIdentityDictionary>>scanFor:. This time > >>is roughly evenly split between primitives and > >>WeakKeyAssociation>>key. Ross, I don't know if this is what you are experiencing but.. I noticed WeakIdentityKeyDictionaries (and WeakValueDictionaries) have major performance issues with #removeKey:, particularly with larger than 3k or so elements. I think it has to do with the 12-bit identityHash which leads to many collisions. I solved the (symptom of) the problem via "MaWeakIdentityKeyDictionary" which simply wraps multiple WeakIdentityDictionaries as necessary to keep them each under 3K elements or so. Believe it or not, this approach is 1000x faster with removeKey: than the standard, and just a little slower on everything else. Its just a temporary (almost three years now!) solution until identityHash can be made better.. - Chris

1 0

Re: Parsing Numbers (was: Re: Exponents of numbers)
by Dave Mason 21 Sep '05

21 Sep '05

Regarding e,q,d as exponents: q and d are used in some worlds to specify extended precision floats, so maybe it makes sense to keep them. the original problem was with radix literals where the radix was greater than 10. I see nothing wrong with 2r110e3 being equal to 48, but 16r1e3 being 4096 is flat-out wrong. Removing d as an exponent designator does nothing to help that. So I see the question of supporting q and d as orthogonal to the radix question. Clearly exponents cannot be used for radices greater than 10, and some might argue that they shouldn't be supported at all for bases other than 10, although I find them convenient. So I suggest fixing the exponent/radix bug, and leaving q and d options as I can't see any damage caused by them. ../Dave

14 20

← Newer
1
...
5
6
7
8
9
10
11
...
27
Older →

Jump to page: