From tismer at stackless.com Mon Mar 15 19:51:47 2004 From: tismer at stackless.com (Christian Tismer) Date: Mon, 15 Mar 2004 19:51:47 +0100 Subject: [Stackless-sprint] Summary In-Reply-To: <40559B97.5050803@gmx.net> References: <40559B97.5050803@gmx.net> Message-ID: <4055FB43.1010304@stackless.com> Lutz Paelike wrote: > I put together a web site right now which sums up the work of each > participant of the sprint. > > Can please everybody sum up shortly (like bob did) what he worked on and > maybe on which topics he will continue to work? > Please post it on the stackless sprint list. So here is what I did: - on the first day I gave a talk which should introduce everybody to Stackless and how it works. With bad preparation, I failed to do a good job on motivating beginners, and Martina dropped off, disappointed. Later on, I could make it a little better, not always diving deeply into internals but showing the current API and the thoughts behind that. - on the second day, I gave a couple of interactive debugger sessions, showing how the interpreter normally works, how it works in Stackless mode, and how switching between tasklets works. I think this was quite informative for everybody. - We had many discussions about the API, and the idea came up that tasklets are even a too large design with too much predefined behavior, and it would be better to find a smaller structure and to implement tasklets on top of it. - Everybody was trying to make some sense out of Stackless, to learn how to use it, and I have as much help as I could. Some concepts were hard to explain in newbie recipes, and we added a few new methods which look more natural to beginners. - I did sime bug fixes and small changes to support the new pre-emptive scheduling, and although I wanted to avoid deep hacking by myself, I wasted almost two days trying to remove the f_callee field and replacing it by a better concept, but failed. - enhanced pre-emtpive scheduling a little by making sure that a tasklet that was in atomic mode while an interrupt occoured gets the delayed interrupt request immediately after it released the atomic flag. One of the seldom changes which work immediately. - Armin asked me many questions about the hows and whys of the implementation, and he asked me astonishingly much about the hardware switching, intial stub, initialization and all the complexity which is introduced by the need to have a main tasklet. - On the last day, he presented me a very small, very clean implementation of "greenlets", which do the hardware switching stuff with the existing library, but implemented as an extension to regular Python! Armin and I had a very long discussion last night, and we found out how to harmonize soft and hard switching in a single concept, which is applicable to any programming language. This is the begining of a completely new implementation of the Stackless kernel, and most probably of a new, general green thread package which is as efficient, clean and small as nothing else before. :-) Things I'm going to do: - continue to take care of Stackless of course, making sure that all the new developers stay happy and they continue to provide more such great documentation, tests, examples, and enhancements. - write documentation, sample code, explain the inner machinery, as it is now and as it should be in the future. - enhance the functionality of the website, to make it easy to upload sample code which should possibly also run on the site, have downloadable installation packages, allow users to add to wiki pages, and more. The following is a list of things to newly develop, but I don't mean to do that alone. Participation is most welcome, this is more a to-do list than my own commitment. - re-implement microthreads based upon tasklets, since people are asking for this. - create a new branch for Stackless on python 2.4, in order to be ready with a Stackless release when Python 2.4 is ready. - create a new branch of Stackless which is the bare minimum to support non-recursive calls, no tasklets, no pickling, minimizing the patches to the Python code base. - create an extension module which does the pickling "from outside", without patching the built-in types in any way. Instead, everything will register itself in the copy_reg module. - create new hard and soft switching based upon the "greenlet" idea. I'm undecided whether to do it as an extension in the first place, but I want to get rid of most of the current machinery. - create tasklets and channels as a Python module, based upon greenlets, for testing and documentation purposes. - implement tasklets as a C extension, based upon greenlets, for compatibility purposes. - all I forgot to mention here should be added, virtually. There are lots of other things to do, especially supporting disemination. My heavy side was always on technical issues, but fortunately I'm no longer alone and can rely on the help and support of the group, which makes me very happy. sincerely -- chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From tismer at stackless.com Thu Mar 18 02:05:03 2004 From: tismer at stackless.com (Christian Tismer) Date: Thu, 18 Mar 2004 02:05:03 +0100 Subject: [Stackless-dev] Design Q. about frames Message-ID: <4058F5BF.10405@stackless.com> Hi friends, I've been hacking quite a bit after the sprint. The state of the source trunk is now almost in a state that this could be the current delivered state. I still didn't change the sprint folder to src, since I'm thinking to do one last change: There is the almost solved issue of f_callee. It *might* be possible to remove it now. The other issue is pickling of running generators. I tried several things, but it turns out that we absolutely need a callback after a generator is run. This can take one of two faces: a) Crate a baseframe object that has the action to be performed. b) Add an f_on_return field to frames which is called on return. Both a) and b) can do the task. Having an extra frame lingering around for every generator call might be doable with some caching. But having the return callback could have some advantages, given that the overhead of an extra call is not too high: - probably, this callback could do the frame deallocation by default and therefore remove f_callee. - this is a cheap callback added to all regular frames, and it can serve more purposes than just generator support, for instance it could be a simple callback that makes __init__ and ohter special methods stackless. All this can be accomplished with an extra frame, of course. But it might make sense to add this simple callback feature, since it would be cheaper. But I'm undecided. Please, give me your input ASAP. On registration for pickling: Both approaches would need to be registered, somehow. b) would involve an extra registry for pickling these callbacks, a) would be simpler since it would go into the f_execute registry. But I don't care, you decide. cheers - chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From lutz_p at gmx.net Mon Mar 15 13:03:35 2004 From: lutz_p at gmx.net (Lutz Paelike) Date: Mon, 15 Mar 2004 13:03:35 +0100 Subject: [Stackless-sprint] Summary Message-ID: <40559B97.5050803@gmx.net> I put together a web site right now which sums up the work of each participant of the sprint. Can please everybody sum up shortly (like bob did) what he worked on and maybe on which topics he will continue to work? Please post it on the stackless sprint list. Thanks, Lutz _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From tismer at stackless.com Thu Mar 18 13:19:57 2004 From: tismer at stackless.com (Christian Tismer) Date: Thu, 18 Mar 2004 13:19:57 +0100 Subject: [Stackless-dev] layout of stackless.com In-Reply-To: <9FA12496-78C4-11D8-A8B4-00039345C610@darwin.in-berlin.de> References: <9FA12496-78C4-11D8-A8B4-00039345C610@darwin.in-berlin.de> Message-ID: <405993ED.2000901@stackless.com> Dinu Gherman wrote: > Stephan Diehl: > >> Then there is still the Zope installation at www.centera.de. Is that >> still >> used? Is somebody moving at least the glossary over? [...] Can we use >> rest >> markup? > > > Well, all this is in a shaky state. I cannot spend a lot of time on > it right now, but I also got burned with the site rebirth. I'd like > to know as well, which Wiki system we have now under Plone and if it > understands ReST? If so, I'll manually carry over the Glossar thing, > which went bust after the big site change. This will also increase > the change for the tutorial thing to take more shape... Plone is a product for Zope. The new site is an additional Zope server with Plone installed. The Wiki is the same ZWiki as in the old site. It is just wrapped by Plone. The tracker is on its own, integration by Zroundup failed, unfortunately. Please tell me of any access problems, and I will try to fix it. My plan is to use the mailman pasword list for all access stuff, but this needs a little time. ciao - chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From tismer at stackless.com Mon Mar 15 18:43:57 2004 From: tismer at stackless.com (Christian Tismer) Date: Mon, 15 Mar 2004 18:43:57 +0100 Subject: [Stackless-sprint] Summary In-Reply-To: <200403151330.20370.stephan.diehl@gmx.net> References: <40559B97.5050803@gmx.net> <200403151330.20370.stephan.diehl@gmx.net> Message-ID: <4055EB5D.70304@stackless.com> Stephan Diehl wrote: >>Can please everybody sum up shortly (like bob did) what he worked on and >>maybe on which topics he will continue to work? Please post it on the >>stackless sprint list. > > > I started on a stacklessness implementation. > Basicly, if it's done, tasklets will be simulated by real threads. > When it's done, it will probably only support a subset of the current > stackless implementation. For example, there won't be (and never will be) any > pickling support. > I'll continue on it later this week. This is just great! > The whole idea of stacklessness might be no longer needed if Armins extension > module (hardswitching) is working. But it will be still very educational. I think all this stuff is extremely helpful, even if it just shows how hard it is to do with threads. I had a longer talk with Armin last evening, and the consequences are quite enormous. I will talk about this in another place. Just this much: His way of hard switching overcomes my old problem that I need a common base address to start with. With his generalized concept, I no longer need this initial stub, I even no longer need any initialization crap, like making sure that I'm "inside" stackless, providing a main tasklet etc. This is all going to be trashed in the future. Even better: I figured out how soft switching fits his new scheme in a very harmonic way. This will be the final cut, best possible solution, *and* it can all be done without a need for Python, it can work with any compiler language. That means: - Stackless gets rid its dependency of hard switching, which might go into an extension (but it might be faster) - Stackless gets rid of pickling support, this can all go into an extension - Stackless gets rid of tasklets, this can be an extension on top of Armin's greenlets - Stackless does nothing else in the kernel than to support non-recursive frame calls. It is a huge storm in my brain, everything is changing, rocks of problems are falling off, the system is going to be very small versatile and elegant. But it's a lot of work, of course. Great that we did this print. Should have done this years before. 1024 thanks to all -- chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From lutz_p at gmx.net Mon Mar 15 20:23:43 2004 From: lutz_p at gmx.net (Lutz Paelike) Date: Mon, 15 Mar 2004 20:23:43 +0100 Subject: [Stackless-sprint] first fprint summary version is online In-Reply-To: <244DCEB8-76B6-11D8-ABEB-00039345C610@darwin.in-berlin.de> References: <244DCEB8-76B6-11D8-ABEB-00039345C610@darwin.in-berlin.de> Message-ID: <405602BF.5080406@gmx.net> > I still only see a dumb "excuse page" for the link above. Also, > I have no account on the new Zope incarnation and my Wiki pages > there are broken, giving me lots of Site Errors. Given all that > I do not really know how to appreciate the move to Plone, al- > though it might not be Plone's fault at all... > > Dinu > Oops. Forgot to include the updated link: http://www.stackless.com/sprint/ repectively http://www.stackless.com/sprint/sprint_participants Lutz _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From tismer at stackless.com Mon Mar 8 21:42:44 2004 From: tismer at stackless.com (Christian Tismer) Date: Mon, 08 Mar 2004 21:42:44 +0100 Subject: [Stackless-sprint] Beamer for Wednesday In-Reply-To: <20040308195758.GE25406@localhost.math.fu-berlin.de> References: <20040308195758.GE25406@localhost.math.fu-berlin.de> Message-ID: <404CDAC4.2030408@stackless.com> Chr. von Stuckrad wrote: > Hi! > > It looks like I got us a beamer for the starting day! > (when I will be with you; having no time on Thursday > and Friday I'll miss the rest...) > > We can get it from the numerics/biocomputing/.. - Workgroup! Suuuuper! Many thanks! ciao - chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From tismer at stackless.com Fri Mar 5 17:37:48 2004 From: tismer at stackless.com (Christian Tismer) Date: Fri, 05 Mar 2004 17:37:48 +0100 Subject: [Stackless-sprint] Sprint Address Message-ID: <4048ACDC.1050207@stackless.com> Dear sprinters, Here is the address of the sprint. Room 025/026 Arnimallee 2-6 14195 Berlin We will meet there on Wednesday, March 10, 10:00 am. Many thanks to Arnold K?hnel, Martin Wilhelm, Christph v. Stuckradt, Dirk Pape and the whole FU staff for making this possible. And here we also have a map (make sure it does not wrap :) : http://www.berlin.de/stadtplan/explorer?ADR_ZIP=14195&ADR_STREET=Arnimallee&ADR_HOUSE=2-6&ADR_INFO=%3Ca+target%3D%22_blank%22+href%3D%22http%3A%2F%2Fwww.stackless.com%22%3EHere+is+the+Stackless+sprint+%28R.025%2F026%29%3C%2Fa%3E ciao - chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From tismer at stackless.com Thu Mar 4 16:54:09 2004 From: tismer at stackless.com (Christian Tismer) Date: Thu, 04 Mar 2004 16:54:09 +0100 Subject: [Stackless-sprint] Bob Ippolito's picture Message-ID: <40475121.4040500@stackless.com> Hi Sprinters, this is basically for me as a reminder whom to pick up on Tuesday. :-) -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From tismer at stackless.com Wed Mar 31 03:56:10 2004 From: tismer at stackless.com (Christian Tismer) Date: Wed, 31 Mar 2004 03:56:10 +0200 Subject: [Stackless-dev] issue tracker Message-ID: <406A253A.3060901@stackless.com> Hi Friends, I see not all developers being included in the issue tracker. At the moment, it appears that bob,peterk,rxe,sdrees,tismer is the current member list. Can you please all become a member, and add as many issues as possible? cheers - chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From bob at redivi.com Tue Mar 2 04:04:29 2004 From: bob at redivi.com (Bob Ippolito) Date: Mon, 1 Mar 2004 22:04:29 -0500 Subject: [Stackless-sprint] Introduction Message-ID: <52ED43D6-6BF6-11D8-93F5-000A95686CD8@redivi.com> Hey, I'm coming to the stackless sprint because I've liked the project for as long as I've known about it, though I've only started actually hacking on and with it recently. My including-but-not-limited-to goals are to help flesh out the new scheduling features, write tests, fix bugs, make sure it works on OS X, meet a bunch of euro python hackers, and hopefully see a little of Berlin. I would probably like to spend a little time rounding off an official "Stackless MacPython" distribution. My Python street cred is mostly limited to the past 2 years of pythonmac-sig and pyobjc-dev, but with a little bit of Twisted, Stackless, pygame, PEAK, and PEP 318 (in no particular order). I stalk the PyPy project, but haven't quite had the time and spare mental capacity to contribute. My blog, "from __future__ import *", seems to be loved by the Daily Python URL lately, so you may recognize me from that even if you don't pay the wonderful Apple Tax. I think I may also be the only US-based non-lurker ;) -bob -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/pkcs7-signature Size: 2357 bytes Desc: not available Url : http://www.stackless.com/pipermail/stackless-dev/attachments/20040301/e7a29131/attachment.bin -------------- next part -------------- _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From tismer at stackless.com Tue Mar 2 03:36:43 2004 From: tismer at stackless.com (Christian Tismer) Date: Tue, 02 Mar 2004 03:36:43 +0100 Subject: [Stackless-sprint] Introductions... In-Reply-To: <4043A9E0.9090903@gmx.net> References: <200403011821.50223.stephan.diehl@gmx.net> <4043A9E0.9090903@gmx.net> Message-ID: <4043F33B.6080704@stackless.com> Lutz Paelike wrote: > Hallo sprinters, > > I'm coming to the sprint too. I don't know yet how i can contribute to > the project but let's see. > I'm living in Berlin so it's not such a big deal to take part. I'm happy to hear that! > I'm sorry i have no spare room and live like Stephan about 1h away from > the event so i guess i'm not the first choice to stay at ;), but in case > somebody is willing to stay together with me in a room it would be > possible at least ... > > Otherwise here is a link to cheap accomodation: > > http://www.ti5.tu-harburg.de/staff/venzke/hostels/#Berlin > > Most of them are located in the center of Berlin. The event takes place > in the west of Berlin. Thank you so much for your kind offer and the hostel information. Questions for the others: Is there a way to get a projector? (I will check with the FU as well) Does somebody know about some video equipment? It was proposed to take records of some sessions and to publish AVI files. I'm not usually keen of such, but in this case I think *everything* that creates publicity will help the project. We must look big, in order to get bigger! next-time-this-will-be-a-conference - ly y'rs -- chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From gherman at darwin.in-berlin.de Mon Mar 15 16:07:54 2004 From: gherman at darwin.in-berlin.de (Dinu Gherman) Date: Mon, 15 Mar 2004 16:07:54 +0100 Subject: [Stackless-sprint] Summary In-Reply-To: <40559B97.5050803@gmx.net> Message-ID: <89384FB9-7692-11D8-8C51-00039345C610@darwin.in-berlin.de> Lutz Paelike: > Can please everybody sum up shortly (like bob did) what he > worked on and maybe on which topics he will continue to work? Ok, here's what I did... - started a "Stackless Glossary" Zwiki on stackless.com - wrote a script to convert it into a compact download-/emailable PDF (needs LaTeX and Docutils, so I'm not sure where to put it - Chris?) - started building a "Stackless Tutorial" thingy with lots of sample code (together with Axel, needs more work) - added a (simple) "umbrella" unittest script to run all unit tests (Stackless/unittests/runAll.py) - suggested API changes like the call-like argument setting thing on tasklets - dito for stackless.getmain/current - started converting some tests to unittests and adding more (e.g. Stackless/unittests/test_chan_cb.py) - asked many questions of varying stupidity - teased my brain with quite a few answers provided by Christian And here's what I'm planning to do... - tell people to jump on extending the "Stackless Glossary" - continue with Axel on a code-rich "Stackless Tutorial" - write something to visualize communication between tasklets - write an article, codename "10 things I learned about Stackless" (feel invited to provide any such items, please!) - add more unittests That's it for today and thanks for the fish! Dinu -- Dinu C. Gherman - http://python.net/~gherman ...................................................................... "The best way to predict the future is to invent it." (Alan Kay) _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From stephan.diehl at gmx.net Mon Mar 15 13:30:20 2004 From: stephan.diehl at gmx.net (Stephan Diehl) Date: Mon, 15 Mar 2004 13:30:20 +0100 Subject: [Stackless-sprint] Summary In-Reply-To: <40559B97.5050803@gmx.net> References: <40559B97.5050803@gmx.net> Message-ID: <200403151330.20370.stephan.diehl@gmx.net> > Can please everybody sum up shortly (like bob did) what he worked on and > maybe on which topics he will continue to work? Please post it on the > stackless sprint list. I started on a stacklessness implementation. Basicly, if it's done, tasklets will be simulated by real threads. When it's done, it will probably only support a subset of the current stackless implementation. For example, there won't be (and never will be) any pickling support. I'll continue on it later this week. The whole idea of stacklessness might be no longer needed if Armins extension module (hardswitching) is working. But it will be still very educational. Stephan _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From tismer at stackless.com Wed Mar 24 18:44:06 2004 From: tismer at stackless.com (Christian Tismer) Date: Wed, 24 Mar 2004 18:44:06 +0100 Subject: [Stackless-dev] sprint trunk renamed to dev Message-ID: <4061C8E6.3060009@stackless.com> Dear former sprinters, I made a copy of the sprint trunk into src and uploaded the current binaries as well. I also renamed the sprint trunk to dev. Sorry, you need to do a new check-out. cheers -- chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From rxe at ukshells.co.uk Tue Mar 16 21:04:00 2004 From: rxe at ukshells.co.uk (Richard Emslie) Date: Tue, 16 Mar 2004 20:04:00 +0000 (GMT) Subject: [Stackless-dev] Summary - Richard Message-ID: Hi Stackless-dev'ers, Here is a brief summary of my sprint checkins: - Added unittest to check life cycle of tasklet during watchdog runs - Added suite of unittests to check watchdog behaved under number of conditions. (Yes the above 2 should be merged - and documented better!) - Hacked uo watchdog to work with soft switching. Some work needs done to work better when pickling (a flag added to pickle state for example) - Demo code to show preemptive autoscheduling with running watchdog functionality. On top of that I mainly tried to absorb an awful lot of information about CPython and stackless implementation - probably not too successfully, but happy that I know an awful more than before. What will I do now? - Seems watchdog broke on final day - so will fix that. Integrate with pickle flag. - Will maintain autoscheduling/watchdog, fix bugs etc if anyone is using this feature in the *current* version. - Possible join in on an effort to implement uthreads library. - Want to strip down to bare minimum stackless and implement global generators. I know Armin has does something similar already, but there is nothing like reimplementing the wheel! - Keep abreast of "greenlets" (nice name :-)) and see if I can add any input there. A lot of good ideas were raised during the sprint and Christian got a little beaten up over things from interface to core implementation. ;-) I guess that's why we have sprints. However, I think everyone should bow lowly and appreciate all the hard work he has put into stackless, which we can now captalise on, and hopefully create great new things. Thankyou Christian. And thanks to anyone who helped organise and make the sprint happen. Cheers, Richard PS Lutz : I don't have a mug shot handy - perhaps we can grab one from the film. PSS Oh and if I owe anyone any money, please write and tell me so I can send you it - otherwise I owe many beers the next at europython. :-) _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From tismer at stackless.com Tue Mar 16 02:36:28 2004 From: tismer at stackless.com (Christian Tismer) Date: Tue, 16 Mar 2004 02:36:28 +0100 Subject: [Stackless-sprint] On greenlets Message-ID: <40565A1C.9030403@stackless.com> Dear friends, first of all, I wish to thank you all for this great sprint!!! I would like to repeat this ASAP, given that I will stand my current monetary disaster. But the sprint was very promising, very productive, and was missing for at least three years. I won't make this mistake again. :-) Also, I'm wondering if we should keep the sprint list and rename it to stackless-dev??? Tomorrow, I will probably purge the 2.3/src trunk and symlink it to 2.3/sprint, but it will be better for you to check it out, again, when this is done. I will keep you posted. Now on the new concept. ----------------------- Armin has given us a very special, wonderful gift, completely unexpected. (Well, not really, if you know Armin...) He implemented greenlets, which are revolutionary better than what I did so far. He didn't know *how* much better they are. Sunday night, we were sitting in the Luise and talked more about it. Armin was not aware of the full power of his implementation, but we figured out that it can be the answer to all questions, namely it is like 42**42. If you want to know more, check out the sprint trunk and read sprint/Greenlet. There is a copy of my/our hard switching platform stuff, and a tiny extension module named greenlet.c, which is able to do hardware switching as an extension module to standard Python!!! We had several sessions at the blackboared, and I tried to explain him how things work. It was especially hard to explain why this initial_stub needs to be recreated on every "entry" to Stackless, it must be a versioned resource, since its history is changing all the time, when Stackless is entered from different contexts. It took him quite some time to understand. We then discussed the implementation more, and I explained how I make sure that all stack slices start on the same top address, and that it is probably possible to go without this restriction, but that my brain was not able to figure out the consequences, and it seemed to be an optimum between simplicity and efficiency, anyway. For the rest of the time, I didn't hear so much from Armin, but some more questions. I appreciated that he was going to swallow all that hackish code to get an understanding. On Sunday evening, he presented me something that changed my life, completely... Now really on the new concept. ------------------------------ What Armin did, is most remarkable. He removed the need to have a common stack base. Instead, he finally found the right model to describe stacks, how they behave and what they are. The breakthrough is to declare that a stack consists of two pieces: One upper part that is really living on the real stack, to some limit, or for the "main" greenlet upto infinity. And there is one lower part which is saved in heap memory. The latter may or may not exist, this is dependant of when other greenlets are created. For further reference, please read and try to understand his implementation. Please don't feel bad if you find it hard to understand. This *is* hard, and I don't believe that there exist so many implementations of this elegance and consequence. There are some crucial, new ideas, and how far the consequences reach, rushed like a ball lightning though my brains. In Armin's design, it is possible to create a greenlet in any context, at any time. There is no need to free up stack space to some initially defined common stack base. Such a thing does not exist. Instead, the machinery always knows which stack patches to save and restore if any greenlet has to be activated, and it does this always right, whatever stack situation might be there. This is the total generalisation of my initial, restricted concept. I'm not jealous because I didn't get this idea in all the years. Instead, I'm more than happy that it has been detected, after all, thanks to Armin. Here the initial advantage that Armin presented to me (but there is more to come): There is no need to create an initial (green|task)let at all. A main greenlet can be described by a static structure. It is the owner of the complete stack, in a way from the topmost possible location (somehow infinite) to the current position. It does not posess a heap based part. Whenever a new greenlet is created, its stack base is defined at the point where we currently are. And this is the great concept: We can create a greenlet wherever we are, there is no need to free up some space occupied by someone else. Later on, greenlets can be activated in any order. The system keeps track of greenlets in chains, which tell which stack pieces need to be saved when activating a greenlet (this only happens when the greenlet to be activated is living above the current one), and which stack pieces have to be restored for this greenlet. Even if the main greenlet is activated and executes some arbitrary code, this concept holds. The initial_stub, which had to be recreated all the time in my concept, is always implicitly associated with the main greenlet, its creation can be avoided, since the main greenlet always *is* this stub, there is no need to create something like a versioned resource. This is simply accomplished by definition. Well, this is hard to understand, but believe me, it is the truth. Now, by this great new input, my brain was reactivated and delivered from former wrong assumptions, and all in a sudden I realised that this was the missing concept over years, which now can get us a very relevant bit further. (And it makes me somewhat happy that Armin didn't see this in the first place, so I don't need to leave the dunce hat on...) How to make greenlets the absolute winner concept ------------------------------------------------- If you look into the implementation, greenlets are created by a g_start function, which defines and claims an area on the stack at the moment the greenlet is created. With a bit of imagination, you will realize that it is not really necessary to claim this initial stack position at the moment the greenlet is created. This can be deferred until the greenlet is actually executed. By doing this, we allow to start a new greenlet regardless from the former stack position, just where it happens to be started. This means, a first activation of a greenlet never needs to cause any stack copying at all. It simply extends the current stack. I'd like to call such a greenlet "placable". For the time the greenlet is executing, it has a claim on the stack. It may decide at some time to jump to a different greenlet. In Armin's first implementation, this will cause some stack saving, at least for the current greenlet. But here the non-recursive scheme chimes into the play: What happens if the greenlet is able to perform a switch without claiming any piece of stack space to be preserved? You guessed it! This is the software switching, integrated into greenlets, seamlessly. If a greenlet is able to jump off without leaving any stack space to be saved, this is a soft switch, and the greenlet stays "placable". This has a *big*, *big* advantage. The idea of placable greenlets says, that we can run soft-switched greenlets on top of any existing stack, we don't need to demand that this stack is "innocent" or have any other restrictions. This was not possible with current Stackless, because I forced "innocent" tasklets to be run at the same zero stack level, all the time. In a sense, this makes placable greenlets as movable and flexible as Python's generators, which also don't care in which context they are called, despite the fact that greenlets are *way* more versatile. So here is my addition to the greenlet concept: - allow greenlets to mark their stack_start when they are *called*, not when they are created. - provide an extra field in the greenlet structure, which points to a memory area where the greenlet can save its current state. - allow the greenlet to jump off after declaring its stack state as being "clean": there is nothing to save, "nothing to declare, I'm innocent". In that case, they *release* their stack_start, staying placable. The consequences are, as far as I can see it right now: - greenlets can always be run for the first time without causing any stack copy operations. - greenlets that have existing heap memory are bound to a specific stack area, and they need to save and restore some stack when they are activated. - greenlets that manage to jump off without leaving state on the stack, but storing it in their extra memory, stay in a placable mode and can be run from any other stack position. - this concept can be implemented without any relationship to the Python language. It may make sense to make it use Python structures, but the concept does not depend on Python at all. - In the case of Python, the extra structure to keep program state off of the C stack is the Python frame or the cframe/baseframe. This could be anything else for other host languages, but the principle remains the same: If you manage to save state in an extra structure, you are not stack based, you can be switched without involving assembly, and it is possible to pickle program state. And how can we make use of all of this? --------------------------------------- I have the deep confirmedness that this new concept will supersede all of what I did with Stackless so far, and that it can be the basis for a general purpose, language independent, most efficient and most versatile implementation of green threads, which has never been designed with this completeness, consequence, elegance, speed and power anywhere and anytime. I am sure that we will be able to create a spin-off of this as a general greenlet package for other compiler languages, and I do hope that we will create *the* best possible green-threading package available, better than everything else, before. Yes, these are big words, but I do believe in it. I think we should use the concept for Python and keep it free software. At the same time, I encourage everybody to think of a commercial spin-off, providing a lightweight task-switching library for Python-less languages and applications, which has no real competitors, and which probably should *not* be free software, but be paid work supporting other non-free software, possibly giving revenues for all of us. Please help to make this take place. I think we are at the bleeding edge here, and we have a great chance if we can do it just right. Apologies about my lengthy brain ejaculation, but I think it was appropriate and necessary. Please let me know about your thoughts, and let me explain things which were hard to understand. enlightened-ly y'rs -- chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From tismer at stackless.com Fri Mar 12 11:33:21 2004 From: tismer at stackless.com (Christian Tismer) Date: Fri, 12 Mar 2004 11:33:21 +0100 Subject: [Stackless-sprint] [Fwd: Re: Stackless crash] Message-ID: <405191F1.8000801@stackless.com> Hi friends, might somebody please try to find this bug? Especially I'd like to know whether this is stackless related or if it is just buggy threading code. thx - chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 pager +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ -------------- next part -------------- An embedded message was scrubbed... From: "Seung Chan Lim" Subject: Re: Stackless crash Date: Fri, 12 Mar 2004 02:00:14 -0500 Size: 38113 Url: http://www.stackless.com/pipermail/stackless-dev/attachments/20040312/f0249c1b/attachment.eml -------------- next part -------------- _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From tmeka at gmx.net Thu Mar 18 17:13:39 2004 From: tmeka at gmx.net (Tomek Meka) Date: Thu, 18 Mar 2004 17:13:39 +0100 (MET) Subject: [Stackless-dev] layout of stackless.com References: <200403181709.40084.stephan.diehl@gmx.net> Message-ID: <31600.1079626419@www61.gmx.net> > Basicly, it's just not linked from anywhere. > Maybe some plone guru could link > http://www.stackless.com/glossar/ > to > http://www.stackless.com/wiki/Cover Well, Dinu just wanted to have glossary separetly from the whole wiki - that's why I've created the glossary folder. What do you think, should we hold the glossar together with wiki or not? To create a link from glossar to wiki is not a big deal... cheers, Tomek _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From bob at redivi.com Fri Mar 19 19:33:34 2004 From: bob at redivi.com (Bob Ippolito) Date: Fri, 19 Mar 2004 13:33:34 -0500 Subject: [Stackless-dev] ODB for Python - possible with stackless In-Reply-To: <0403191015380.-3@example.com> References: <0403191015380.-3@example.com> Message-ID: Yes, you could even unpickle it on another architecture entirely. Some pickles are not resumable (if you pickle something that requires state from an interrupted written-in-C function), but the object graph would be perfectly good for this kind of debugging. On Mar 19, 2004, at 1:17 PM, seanj wrote: > Will I also be able to pickle running code and unpickle it on another > machine > to resume execution? > > If this is mentioned somewhere please point me to it, I don't follow > this list > very actively. > > On Fri, 19 Mar 2004, Bob Ippolito wrote: > >> The pickling/marshalling-of-anything combined with autoscheduling or >> appropriately placed schedule() calls would allow us to develop >> something like this for Python: >> >> http://www.lambdacs.com/debugger/debugger.html _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From tismer at stackless.com Thu Mar 18 13:39:47 2004 From: tismer at stackless.com (Christian Tismer) Date: Thu, 18 Mar 2004 13:39:47 +0100 Subject: [Stackless-dev] layout of stackless.com In-Reply-To: <405993ED.2000901@stackless.com> References: <9FA12496-78C4-11D8-A8B4-00039345C610@darwin.in-berlin.de> <405993ED.2000901@stackless.com> Message-ID: <40599893.3030502@stackless.com> Christian Tismer wrote: Ok, some additions. > Dinu Gherman wrote: ... >> Well, all this is in a shaky state. I cannot spend a lot of time on >> it right now, but I also got burned with the site rebirth. I'd like >> to know as well, which Wiki system we have now under Plone and if it >> understands ReST? If so, I'll manually carry over the Glossar thing, >> which went bust after the big site change. This will also increase >> the change for the tutorial thing to take more shape... Plone understands DynamicDocument now, which means page templates. The Wiki understands ReST. I had created documents for everybody in their folders, but figured out that this gives wrong ownership. Not nice, but this helped: Now every sprint member has both the member and the manager role, which means they can change *everything*. This is a bit of a risk and might be fine-tuned, later. But I'm pretty sure that we all can edit our pages, now :-) -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From stephan.diehl at gmx.net Mon Mar 1 18:21:50 2004 From: stephan.diehl at gmx.net (Stephan Diehl) Date: Mon, 1 Mar 2004 18:21:50 +0100 Subject: [Stackless-sprint] Introductions... Message-ID: <200403011821.50223.stephan.diehl@gmx.net> Hallo sprinters, I don't know yet for how long I can attend the sprint (bad timing). Nonetheless, since I'm living here in Berlin, one person could sleep in my place. But be aware, that it takes quite a long time from my place to the University (about 1h with public transport). About my bad timing: I'm singing in a choir and we have a concert on sunday, the 14th (5pm). Bach: Johannespassion (St. John passion) If somebody is interested, I could organize some tickets (not for free). _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From gherman at darwin.in-berlin.de Mon Mar 8 09:54:05 2004 From: gherman at darwin.in-berlin.de (Dinu Gherman) Date: Mon, 8 Mar 2004 09:54:05 +0100 Subject: [Stackless-sprint] Introductions... In-Reply-To: <4040C9C0.8010509@stackless.com> Message-ID: <27AFA63A-70DE-11D8-99A5-00039345C610@darwin.in-berlin.de> Christian Tismer: > May I also ask everybody to give a brief introduction of > himself, where he comes from, and what his expectations > for the sprint are? Just a few lines, and real names > are better than the email acronyms. Sorry for dropping in "just in time"... I'll join the Stackless sprint team for a variety of reasons, the most prominent ones being: - Python needs better support for low-level parallelism - Stackless needs more developpers (docs, tools, demos) - Stackless needs more mindshare (docs, articles, etc) - Christian needs more helping hands - I need a better understanding of Stackless - it is high time for the Sprint to happen - it's around my corner, too ;-) So far, I've been barely getting my feet wet with Stackless, compiling it on OS X and playing with the demos, but I'd love to help with developping more demos and tests and pulling infor- mation out of the core developpers that might end up in some kind of documentation and articles to increase the Stackless mindshare. Eventually, I'd like to use Stackless more tightly combined with PyObjC on OS X. Expectations, well... I hope I'll gain a better understanding at the sprint of the internals and concepts of Stackless and a variety of use cases for it - ideally to a degree that I can develop my own demos illustrating Stackless features and write articles about it, which is a way of repeating and deepening what I've learned. See you soon... Dinu -- Dinu C. Gherman - http://python.net/~gherman ...................................................................... "I once asked Ivan, 'How is it possible for you to have invented computer graphics, done the first object oriented software system and the first real time constraint solver all by yourself in one year?' And he said 'I didn't know it was hard.'" (Alan Kay on Ivan Sutherland) _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From tismer at stackless.com Mon Mar 15 20:50:06 2004 From: tismer at stackless.com (Christian Tismer) Date: Mon, 15 Mar 2004 20:50:06 +0100 Subject: [Stackless-sprint] first fprint summary version is online In-Reply-To: <405602BF.5080406@gmx.net> References: <244DCEB8-76B6-11D8-ABEB-00039345C610@darwin.in-berlin.de> <405602BF.5080406@gmx.net> Message-ID: <405608EE.8010703@stackless.com> Lutz Paelike wrote: > >> I still only see a dumb "excuse page" for the link above. Also, >> I have no account on the new Zope incarnation and my Wiki pages >> there are broken, giving me lots of Site Errors. Given all that >> I do not really know how to appreciate the move to Plone, al- >> though it might not be Plone's fault at all... >> >> Dinu >> > Oops. Forgot to include the updated link: > > http://www.stackless.com/sprint/ > > repectively > > http://www.stackless.com/sprint/sprint_participants Yup, I found out, meanwhile, and edited things a little bit. I like this very very much!! At the moment, this is a big HTML page in one shot, fine so far. What I think might make sense is to: Split it into pices, by-person, merging stuff together by turing the main page into a TAL page which expands a little page from each person't home. Is that possible or are there access restrictions? I can do this if I know where things should go. Could be, for instance, every user page just defines a macro, and the main pages uses metal:use-macro to fill the table with text. ciao - chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From tismer at stackless.com Thu Mar 18 01:54:48 2004 From: tismer at stackless.com (Christian Tismer) Date: Thu, 18 Mar 2004 01:54:48 +0100 Subject: [Stackless-dev] Sprint Page updated Message-ID: <4058F358.2050202@stackless.com> Dear Sprinters, The Sprint Participants page was changed. It does no longer contain your text, but references to documents which I created in your home folders. If you want to update or fill with text, finally (Armin, Tomek), please edit the page which I placed into your home. You can change everything but the enclosing span tag, which defines a macro that is referred to in the assembled page. Thanks for all the fish -- chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From tismer at stackless.com Tue Mar 2 04:24:32 2004 From: tismer at stackless.com (Christian Tismer) Date: Tue, 02 Mar 2004 04:24:32 +0100 Subject: [Stackless-sprint] Introduction In-Reply-To: <52ED43D6-6BF6-11D8-93F5-000A95686CD8@redivi.com> References: <52ED43D6-6BF6-11D8-93F5-000A95686CD8@redivi.com> Message-ID: <4043FE70.4000205@stackless.com> Bob Ippolito wrote: [great! to have you here] > I think I may also be the only US-based non-lurker ;) Stackless was heavily sponsored by IronPort, San Bruno CA, for 17 months, and I tried to convince them to send Sam Rushing over. If somebody should happen *not* to know him: he wrote Asyncore and Medusa, still crucial for today's Zope. And he is guilty for starting the Stackless project, since he said "one cannot do that" on python-dev. Unfortunately, his company cannot afford to send him off, which makes me sad, mainly about the company. But I'm sure you will keep us *very* busy. -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From gherman at darwin.in-berlin.de Thu Mar 18 11:11:29 2004 From: gherman at darwin.in-berlin.de (Dinu Gherman) Date: Thu, 18 Mar 2004 11:11:29 +0100 Subject: [Stackless-dev] layout of stackless.com In-Reply-To: <200403181048.53077.stephan.diehl@gmx.net> Message-ID: <9FA12496-78C4-11D8-A8B4-00039345C610@darwin.in-berlin.de> Stephan Diehl: > Then there is still the Zope installation at www.centera.de. Is that > still > used? Is somebody moving at least the glossary over? [...] Can we use > rest > markup? Well, all this is in a shaky state. I cannot spend a lot of time on it right now, but I also got burned with the site rebirth. I'd like to know as well, which Wiki system we have now under Plone and if it understands ReST? If so, I'll manually carry over the Glossar thing, which went bust after the big site change. This will also increase the change for the tutorial thing to take more shape... Dinu _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From tismer at stackless.com Fri Mar 19 19:31:10 2004 From: tismer at stackless.com (Christian Tismer) Date: Fri, 19 Mar 2004 19:31:10 +0100 Subject: [Stackless-dev] ODB for Python - possible with stackless In-Reply-To: <0403191015380.-3@example.com> References: <0403191015380.-3@example.com> Message-ID: <405B3C6E.9080504@stackless.com> seanj wrote: > Will I also be able to pickle running code and unpickle it on another machine > to resume execution? Yes. As far as your code is running top-level, that means, there is no extra C state, your pickle can be run on any other machine. cheers - chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From gherman at darwin.in-berlin.de Wed Mar 31 09:38:00 2004 From: gherman at darwin.in-berlin.de (Dinu Gherman) Date: Wed, 31 Mar 2004 09:38:00 +0200 Subject: [Stackless-dev] ViewCVS, RSS Message-ID: <564FD4EC-82E6-11D8-9D4E-00039345C610@darwin.in-berlin.de> Hi, would anybody like to setup ViewCVS for the Stackless CVS repository? Christian is too busy, for good reasons. If we had it I would have a good reason to undust a projectlet of mine, cvs2rdf which monitors a CVS/SVN repository and generates RSS files for the increasingly popular RSS readers with links pointing into a ViewCVS website. I'd then try adding CVS/SVN checkin messages as well... Here are some samples I'm creating automatically for about two years now for varying time intervals, usually 24-72 hours, updated every hour: Python: http://me.in-berlin.de/~darwin/cvspysf.rdf PyObjC: http://me.in-berlin.de/~darwin/cvspyobjcsf.rdf PyPy: http://me.in-berlin.de/~darwin/svnpypycs.rdf ReportLab: http://me.in-berlin.de/~darwin/cvsrlsf.rdf Regards, Dinu -- Dinu C. Gherman - http://python.net/~gherman ...................................................................... "The best way to predict the future is to invent it." (Alan Kay) _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From seanj at xyke.com Fri Mar 19 19:17:50 2004 From: seanj at xyke.com (seanj) Date: Fri, 19 Mar 2004 10:17:50 -0800 (PST) Subject: [Stackless-dev] ODB for Python - possible with stackless In-Reply-To: References: Message-ID: <0403191015380.-3@example.com> Will I also be able to pickle running code and unpickle it on another machine to resume execution? If this is mentioned somewhere please point me to it, I don't follow this list very actively. Sean On Fri, 19 Mar 2004, Bob Ippolito wrote: > The pickling/marshalling-of-anything combined with autoscheduling or > appropriately placed schedule() calls would allow us to develop > something like this for Python: > > http://www.lambdacs.com/debugger/debugger.html > > -bob > > > _______________________________________________ > Stackless-dev mailing list > Stackless-dev at stackless.com > http://www.stackless.com/mailman/listinfo/stackless-dev > _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From A.Bringenberg at srz-berlin.de Mon Mar 15 19:27:36 2004 From: A.Bringenberg at srz-berlin.de (Axel Bringenberg) Date: Mon, 15 Mar 2004 19:27:36 +0100 Subject: [Stackless-sprint] Summary In-Reply-To: <40559B97.5050803@gmx.net> References: <40559B97.5050803@gmx.net> Message-ID: <4055F598.1050807@srz-berlin.de> Lutz Paelike schrieb: [...] > Can please everybody sum up shortly (like bob did) what he worked on and > maybe on which topics he will continue to work? > Please post it on the stackless sprint list. ... and here is my little summary: I joined the sprint as a stackless newbie dummy, tried to figure out the way (with all these mines and pitfalls) of a normal stackless beginner. In this role I created a lot of ten-lines-examples, the base of a didactic step-by-step beginners tutorial. And in future? Well, there's a lack of documentation which shouldn't be ignored ;-) Thanks for all, - bringi _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From tismer at stackless.com Fri Mar 26 04:58:17 2004 From: tismer at stackless.com (Christian Tismer) Date: Fri, 26 Mar 2004 04:58:17 +0100 Subject: [Stackless-dev] New Stacjless Protocol Message-ID: <4063AA59.6020803@stackless.com> New Stackless Protocol ---------------------- Stackless has used some extra fields in the type objects, to support non-recursive calls. After dynamic classes evolved more and more, keeping track of such extra fields has become more and more difficult. In extent, I had to maintain quite a large number of function copies, which behaved slightly different. After all, this approach didn't give me all the results I expected. Since yesterday, I have been working intensively on a new approach. It might be sligtly less efficient, but it seems to be very, very clean. tp_call_nr and tp_iternext_nr have been removed from type objects. So far, type objects are unchanged, again. Stackless instead now supports an implicit parameter for every function call. It would be great if we could make it into an explicit parametzer, but that would change the implementation completely. Information about the concrete implementation should probably go somewhere else. Just summarizing: Stackless passes all Python tests and all of its internal tests, the code base is reduced, again, and we have a couple of noew non-recursive calls: - tasklet.raise_exception can run stackless, - tasklet.kill can run stackless, - generator.next can run stackless, like "for i in generator" could already, last time. This version of stackless is not accessible via the stackless cvs. If you want to try it, you need to use cvs co slpdev/src/2.3/dev Will add a link to this to the public pages, soon. cheers -- chris p.s.: please look into the stackless news items. p.p.s.: I'm eagerly looking for new activities of the former sprinters. Are you missing something from me??? -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From rxe at ukshells.co.uk Mon Mar 1 17:37:10 2004 From: rxe at ukshells.co.uk (Richard Emslie) Date: Mon, 1 Mar 2004 16:37:10 +0000 (GMT) Subject: [Stackless-sprint] Introductions... Message-ID: Hi fellow sprinters, I am coming to sprint! A few months after euro python, I had a free weekend and was wondering how to enable autoscheduling in stackless, only to find it hadn't been implemented yet. To discover this I ended up browsing the code to make sure I hadn't missed an obvious flag. Giving up on finding a flag, I realised a quick hack that would do what I wanted - so an hours hacking later I had autoscheduling stackless. Simply this did internal stackless schedule calls every n bytecodes for some configurable n - all the hooks were there so I can't claim much credit. Since then the internals have changed somewhat, and autoscheduling hasn't kept in sync... so here's to doing autoscheduling properly second time around. :-) Anyways that's my introduction and I was at the PyPy sprint in Amsterdam, so some might know me through that. :-) See you there & cheers, Richard PS I need an address v. soon, so I can make travel arrangements etc.... thanks. _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From tismer at stackless.com Sat Mar 6 18:39:29 2004 From: tismer at stackless.com (Christian Tismer) Date: Sat, 06 Mar 2004 18:39:29 +0100 Subject: [Stackless-sprint] Introductions... In-Reply-To: References: Message-ID: <404A0CD1.2030107@stackless.com> Richard Emslie wrote: > Hi fellow sprinters, > > I am coming to sprint! Hi Richard, did you send us your itinerary, already? I need to knwo when and where you arrive, in order to pick you up. Here is what I have so far: Bob Ippolito: Berlin, Germany Tegel (TXL) Tue, Mar 09 at 12:10pm I will try to pick him up in Tegel and bring him home since he will probably be terribly jet-laged and needs a nap. Armin Rigo: Tuesday 9 at 17:15 at Berlin Sch?nefeld. Armin is residing in Stephan Diehl's flat, so I assume that you will take care of him, Stephan? Richard Emslie: No arrival information, yet. Richard and Bob stay in my house in Steinst?cken. Maybe we all want to meet there, on Tuesday evening? ciao - chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From tismer at stackless.com Tue Mar 2 05:48:31 2004 From: tismer at stackless.com (Christian Tismer) Date: Tue, 02 Mar 2004 05:48:31 +0100 Subject: [Stackless-sprint] Intruduction Message-ID: <4044121F.6000209@stackless.com> """ Please allow me to introduce myself, I'm a man of wealth and taste I've been around for a long, long year, Stole many a man's soul and faith And I was 'round when Jesus Christ, Had his moment of doubt and pain Made damn sure that Pilate, Washed his hands and sealed his fate CHORUS: Pleased to meet you Hope you guess my name But what's puzzling you Is the nature of my game """ (1) This was Stackless 1.0, continuations and blasphemy, starting 1999. Yes, I mean the song, I think it really applied to me. Stackless was an extremely difficult approach, since I supported full continuations. (Will give an extra session on that if you like). Short after that, Stackless Microthreads were created. Funnily, CCPgames wrote their first version of the EVE online game on top of microthreads. In those days, I felt I had to support continuations, as the strongest possible concept that IS had ever created. This is still true, but I also did the biggest harm to Stackless, since Guido hates continuations. If I had had an approach like today, chances for some integration would have been much better. But I was greedy and snotnosed. Then, Stackless fell into some agony for a while. In 2002, I started over, with a rigorous new 2.0 design. This happened in a cafe, after cake and three cans of coffee. Much simpler, brute-force hardware stack fiddling. Just great for porting, a new Python version took < one day. I went to Rejkjavik and ported the CCPgames EVE application to Stackless 2.0, with great success - things were about four times faster. This was also the birth of tasklets and channels, both concepts were stolen from Limbo, with its attractive simplicity. At the same time, I got a contract with a US firm to support them. After I failed to port their application, I started to merge 1.0 features back into Stackless, in late 2002. Later on they claimed this had been my primary task, but this is not true. This turned out to become Stackless 3.0, and unfortunately, they never adopted it, but began to write their own minimalist Stackless Python. Not to mention a small contract with TwinSun.com for thread pickling. This has become a major application of Stackless, today. It also has become a major porting problem today, as I feared it: Pickling is part of Stackless, but newer Pythons grow more and more objects which need pickling, and this costs a lot of resources. But I love it, anyway. ---- Enough of that history stuff. I have to tell you something crucial about my person. Well, there was one continous thread throughout my life: When I was a boy, I loved to work on my toy railway. This lasted for many years, and my biggest fear was to get this thing ready, at one day. This has been one of my problems with Stackless, too. The result: - Almost no documentation - special methods which nobody can understand - almost no tests (well, Bob did a *good* job meanwhile) - very few examples - it never gets really ready, upto now As a summary, Stackless has been a very egocentric project, making it into an exception that a company adopted the technology, although it is said to be really useful. But who would take the risk? By this sprint, I'm trying to get rid of my toy railway behavior. As a friend likes to say: "Try to get mature before becoming senile". I want to share what I know, I want to get help with documentation and examples, and I want to create a broader basis of people who know about Stackless and how it works. After all, it is not that hard! Just a bunch of simple ideas and their consequences for the coding. I very much hope this sprint will create the critical mass to make more companies take the risk of using Stackless Python. Finally, I want to start an Alliance of supporters, which makes companies confident to join the club. They should join by spending money or by human resources. Make a Stackless basis of five people, and we are all set. This is what it needs. Now the sprint seems to make a basis of more than ten people, which is even better! Ok, and finally I would love to see a subset of Stackless to go into the Python kernel. HaaHaa, good joke :-)) Happy to see you all next Wednesday -- chris (1) The Rolling Stones - Sympathy For The Devil -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From tismer at stackless.com Fri Mar 26 19:39:33 2004 From: tismer at stackless.com (Christian Tismer) Date: Fri, 26 Mar 2004 19:39:33 +0100 Subject: [Stackless-dev] Re: channel object becoming PYCObject or long? In-Reply-To: <00d601c41361$58394440$6401a8c0@djslim> References: <068a01c412d9$9fc55750$95fe46c0@GIRLS> <4063A240.8060600@stackless.com> <002101c412e3$978ccdc0$6401a8c0@djslim> <4064588E.8080306@stackless.com> <5BEE900A-7F49-11D8-8824-000A95686CD8@redivi.com> <00d601c41361$58394440$6401a8c0@djslim> Message-ID: <406478E5.90709@stackless.com> Seung Chan Lim wrote: > Has anyone seen your channel object all of a sudden turn into a PyCObject or > a long object? > > This is very bizarre, and I'm trying to get it to happen consistently, but > having some trouble. Thougt I'd see if anyone else has seen this happen Sounds like a refcount problem during thread destruction. This should definately not happen. ciao - chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From stucki at math.fu-berlin.de Tue Mar 9 18:26:45 2004 From: stucki at math.fu-berlin.de (Chr. von Stuckrad) Date: Tue, 9 Mar 2004 18:26:45 +0100 Subject: [Stackless-sprint] Meeting at Christian Timer's after 20:00 (8pm ff) Message-ID: <20040309172645.GK25406@localhost.math.fu-berlin.de> Hi! I've just been asked to forward a reminder for the casual meeting at: Christin Tismer's (Home) TODAY this evening after 20:00 (local time :-) See the exact address on the signature of every mail he sent to the list! Sincerely yours Stucki (postmaster where the sprint will be held) -- Christoph von Stuckrad * * |nickname |\ Freie Universitaet Berlin |/_*|'stucki' |Tel(days):+49 30 838-75 459| Fachbereich Mathematik, EDV|\ *|if online|Tel(else):+49 30 77 39 6600| Arnimallee 2-6/14195 Berlin* * |on IRCnet|Fax(alle):+49 30 838-75454/ _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From tismer at stackless.com Sun Mar 14 12:43:25 2004 From: tismer at stackless.com (Christian Tismer) Date: Sun, 14 Mar 2004 12:43:25 +0100 Subject: [Stackless-sprint] Re: [Stackless] Warning: build may fail on OS X with Xcode 1.1.1's gcc 3.3 compiler In-Reply-To: <475F8E50-74E5-11D8-B71A-000A95686CD8@redivi.com> References: <475F8E50-74E5-11D8-B71A-000A95686CD8@redivi.com> Message-ID: <4054455D.3050204@stackless.com> Bob Ippolito wrote: > gcc version 3.3 20030304 (Apple Computer, Inc. build 1636) (-mlongcall), > distributed with the latest seed of Xcode 1.1.1, isn't able to build > slp_transfer. You will see errors that looks like this: > > /var/tmp//ccyq4M9K.s:64:non-relocatable subtraction expression, "__cst" > minus "L00000000001$pb" > /var/tmp//ccyq4M9K.s:64:symbol: "L00000000001$pb" can't be undefined in > a subtraction expression > > Current workarounds: > don't upgrade Xcode > or compile with gcc 3.1 (/usr/bin/gcc3) > > If I can't tweak it to compile, I'll guess I'll distribute a .s > (assembly) file for the platform (created with gcc -S from a working > compiler) and jigger the makefile to use that instead. That's just fine. Maybe this would be the better option in general, since we would have more control over the registers and such. On the other hand, it would not integrate so nice with those macros which do the actual data transfers. Maybe these would be real functions, instead, and called from the assembly? -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 pager +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From lutz_p at gmx.net Mon Mar 1 22:23:44 2004 From: lutz_p at gmx.net (Lutz Paelike) Date: Mon, 01 Mar 2004 22:23:44 +0100 Subject: [Stackless-sprint] Introductions... In-Reply-To: <200403011821.50223.stephan.diehl@gmx.net> References: <200403011821.50223.stephan.diehl@gmx.net> Message-ID: <4043A9E0.9090903@gmx.net> Hallo sprinters, I'm coming to the sprint too. I don't know yet how i can contribute to the project but let's see. I'm living in Berlin so it's not such a big deal to take part. I'm sorry i have no spare room and live like Stephan about 1h away from the event so i guess i'm not the first choice to stay at ;), but in case somebody is willing to stay together with me in a room it would be possible at least ... Otherwise here is a link to cheap accomodation: http://www.ti5.tu-harburg.de/staff/venzke/hostels/#Berlin Most of them are located in the center of Berlin. The event takes place in the west of Berlin. Cheers, Lutz Paelike _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From bob at redivi.com Fri Mar 19 17:24:31 2004 From: bob at redivi.com (Bob Ippolito) Date: Fri, 19 Mar 2004 11:24:31 -0500 Subject: [Stackless-dev] ODB for Python - possible with stackless Message-ID: The pickling/marshalling-of-anything combined with autoscheduling or appropriately placed schedule() calls would allow us to develop something like this for Python: http://www.lambdacs.com/debugger/debugger.html -bob _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From stephan.diehl at gmx.net Thu Mar 18 17:09:40 2004 From: stephan.diehl at gmx.net (Stephan Diehl) Date: Thu, 18 Mar 2004 17:09:40 +0100 Subject: [Stackless-dev] layout of stackless.com In-Reply-To: <9FA12496-78C4-11D8-A8B4-00039345C610@darwin.in-berlin.de> References: <9FA12496-78C4-11D8-A8B4-00039345C610@darwin.in-berlin.de> Message-ID: <200403181709.40084.stephan.diehl@gmx.net> > [...] If so, I'll manually carry over the Glossar thing, > which went bust after the big site change. This will also increase > the change for the tutorial thing to take more shape... I've found the Glossary. It's not where one would expect it, but under http://www.stackless.com/wiki/FrontPage/contents Basicly, it's just not linked from anywhere. Maybe some plone guru could link http://www.stackless.com/glossar/ to http://www.stackless.com/wiki/Cover cheers Stephan > > Dinu > > > _______________________________________________ > Stackless-dev mailing list > Stackless-dev at stackless.com > http://www.stackless.com/mailman/listinfo/stackless-dev _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From tismer at stackless.com Sat Mar 6 20:07:14 2004 From: tismer at stackless.com (Christian Tismer) Date: Sat, 06 Mar 2004 20:07:14 +0100 Subject: [Stackless-sprint] Does somebody have a projector for a day or two? Message-ID: <404A2162.6030604@stackless.com> Hi Sprinters, I asked the FU staff whether we can get a projector. They said this is most probably not possible. Does somebody of the Berlin-Sprinters have access to a projector from his company, maybe? It would be nice to have one, at least for Wednesday. Although I also can create a file for download, and then we need to use our laptops. A projector would in fact be nicer, if we have that option. If you can manage this, please tell it to this list, since we don't need three of them. :-) ciao - chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From stephan.diehl at gmx.net Thu Mar 18 17:24:53 2004 From: stephan.diehl at gmx.net (Stephan Diehl) Date: Thu, 18 Mar 2004 17:24:53 +0100 Subject: [Stackless-dev] layout of stackless.com In-Reply-To: <31600.1079626419@www61.gmx.net> References: <200403181709.40084.stephan.diehl@gmx.net> <31600.1079626419@www61.gmx.net> Message-ID: <200403181724.53982.stephan.diehl@gmx.net> On Thursday 18 March 2004 17:13, you wrote: > > Basicly, it's just not linked from anywhere. > > Maybe some plone guru could link > > http://www.stackless.com/glossar/ > > to > > http://www.stackless.com/wiki/Cover > > Well, Dinu just wanted to have glossary separetly from the whole wiki - > that's why I've created the glossary folder. What do you think, should we > hold the glossar together with wiki or not? To create a link from glossar > to wiki is not a big deal... As long as they are not interlinked, they are seperate, aren't they? I'll try to figure out the linking (have to learn it anyway :-) Stephan > > cheers, > Tomek _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From senn at maya.com Tue Mar 2 18:09:18 2004 From: senn at maya.com (Jeff Senn) Date: Tue, 2 Mar 2004 12:09:18 -0500 Subject: [Stackless-sprint] Introduction In-Reply-To: <52ED43D6-6BF6-11D8-93F5-000A95686CD8@redivi.com> References: <52ED43D6-6BF6-11D8-93F5-000A95686CD8@redivi.com> Message-ID: <57118F9E-6C6C-11D8-B5C4-000393C94CC4@maya.com> On Mar 1, 2004, at 10:04 PM, Bob Ippolito wrote: > I think I may also be the only US-based non-lurker ;) Well...I've been lurking lately. I cannot make the sprint, but am very interested in autoscheduling as I've been working on a research prototype that makes use of an autoscheduling-enabled version of Stackless 2.0 (and Stackless 1.0 before that!) since the beginning. So, unless anyone minds my eavesdropping, I'd like to monitor the sprint discussion... -Jas - - - - - - - - - - - - - - / / |-/ \ / /| ? Jeff Senn MAYA Design /|/| |/ o | /-| Technologist Pittsburgh, PA Taming Complexity ? senn at maya.com www.maya.com _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From lutz_p at gmx.net Mon Mar 15 17:40:02 2004 From: lutz_p at gmx.net (Lutz Paelike) Date: Mon, 15 Mar 2004 17:40:02 +0100 Subject: [Stackless-sprint] first fprint summary version is online Message-ID: <4055DC62.5020302@gmx.net> A first version of a sprint summary is available: http://www.stackless.com/Members/lutz/SDocument.2004-03-15.0586225618/sdocument_view Still missing: Summary of Christian, Armin, Axel, Stefan, Richard, Tomek Richard we forgot to take a portrait picture of you. Can you send me one ? By the way if somebody wants to change/remove his picture you can mail me. Tomek what's your last name ? Cheers, Lutz _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From gherman at darwin.in-berlin.de Mon Mar 15 20:22:47 2004 From: gherman at darwin.in-berlin.de (Dinu Gherman) Date: Mon, 15 Mar 2004 20:22:47 +0100 Subject: [Stackless-sprint] first fprint summary version is online In-Reply-To: <24291.1079376390@www52.gmx.net> Message-ID: <244DCEB8-76B6-11D8-ABEB-00039345C610@darwin.in-berlin.de> Tomek Meka: > http://www.stackless.com/Members/lutz/SDocument.2004-03-15.0586225618/ > sdocument_view > > I've forgotten yesterday to change to default state for folders - and > thus > were the images in lutz's folder not readable for all. I've changed > it, it > now > I do not experience any problems with viewing the page, also as an > anonymous > user. I still only see a dumb "excuse page" for the link above. Also, I have no account on the new Zope incarnation and my Wiki pages there are broken, giving me lots of Site Errors. Given all that I do not really know how to appreciate the move to Plone, al- though it might not be Plone's fault at all... Dinu _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From arigo at tunes.org Mon Mar 1 18:25:02 2004 From: arigo at tunes.org (Armin Rigo) Date: Mon, 1 Mar 2004 17:25:02 +0000 Subject: [Stackless-sprint] Introduction Message-ID: <20040301172502.GA393@vicky.ecs.soton.ac.uk> Hello, I'm coming to Christian's sprint for the fun of C hacking, as I used to do while hacking on Psyco :-) BTW I need some accomodation. So in answer to Christian's suggestion to ask if someone has room in his flat, let me ask: would someone have room in his flat? Otherwise I welcome cheap hotel recommendations (even just helping me book one would be excellent, if it doesn't have on-line booking). I'm arriving Tuesday 9 at 17:15 at Berlin Schonefeld, and leaving Monday 15 from there at 9:40. Looking forward to see you all, Armin. _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From tismer at stackless.com Tue Mar 16 16:07:43 2004 From: tismer at stackless.com (Christian Tismer) Date: Tue, 16 Mar 2004 16:07:43 +0100 Subject: [Stackless-dev] The sprint list has been renamed to dev Message-ID: <4057183F.1060809@stackless.com> Dear former Sprinters, you now have become developers, and therefore the list was renamed to stackless-dev. Please make sure to adjust the list address if you happen to reply to an older message. cheers - chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From lutz_p at gmx.net Mon Mar 15 21:09:18 2004 From: lutz_p at gmx.net (Lutz Paelike) Date: Mon, 15 Mar 2004 21:09:18 +0100 Subject: [Stackless-sprint] first fprint summary version is online In-Reply-To: <405608EE.8010703@stackless.com> References: <244DCEB8-76B6-11D8-ABEB-00039345C610@darwin.in-berlin.de> <405602BF.5080406@gmx.net> <405608EE.8010703@stackless.com> Message-ID: <40560D6E.2080107@gmx.net> > At the moment, this is a big HTML page in one shot, fine > so far. What I think might make sense is to: > Split it into pices, by-person, merging stuff together > by turing the main page into a TAL page which expands > a little page from each person't home. I agree. It would be nice to keep the content dynamic. However i was thinking of a static "snapshot" of the sprint results. Work which is continued upon the sprint results in the future should not be kept in the sprint folder but in a better place. Right now everything is just quickly hacked together. > Is that possible or are there access restrictions? > I can do this if I know where things should go. > Could be, for instance, every user page just defines > a macro, and the main pages uses metal:use-macro > to fill the table with text. yes that would be the way to go probably. i don't see how you can do that through the plone interface. i tried to create a zpt page in the zmi but that does not show up in the plone interface because it is not a cmf object. a direct link to a zpt should work though. Ideas? Tomek? Lutz _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From sdrees at sdrees.de Mon Mar 15 23:24:48 2004 From: sdrees at sdrees.de (Stefan Drees) Date: Mon, 15 Mar 2004 23:24:48 +0100 Subject: [Stackless-sprint] Summary sdrees Message-ID: <20040315222448.GA22442@knoten.biz> Dear stackless, # ----- 8< ----------------------------- my sprint-summary: ================== * Always a great time, if someone opens up and a team is willing to close on in (attention sloppy language usage;) or as some of our kids say _reeeespect_ chris. * Count me in for the next _Zwischensprint_ in Berlin! * Came late friday evening, had to perform some extra steps to sync with the team's current tasks * Came closer to the point with stackless, so to be able in helping in not letting the core ideas become pointless. * Began to wade through the C-code. * Performed clean-up and patching tasks where needed. * Discussed test and validation strategies with Dinu and Chris. * Patched some python core tests (generator, descr) that tested for some silly conditions, failing with stackless. my future plans: ================ * Enhance funding for stackless. Hm, can *you* do this to ;?). * Help * hacking the C-part of it and * help with the greenlets and * keeping the linux2 plattform working and * enhance testing (function and performance) * Contribute to Wiki and Cookbook on stackless.com * Stay tuned ;) All the best, Stefan. # ----- 8< ------ Best regards / MfG, Stefan Drees. -- .o. e-mail: stefan at drees.name, web: www.sdrees.org ..o fingerprint = 516C C4EF 712A B26F 15C9 C7B7 5651 6964 D508 1B56 ooo stefan drees - consulting and lecturing - problems to tasks _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From tismer at stackless.com Mon Mar 15 19:39:47 2004 From: tismer at stackless.com (Christian Tismer) Date: Mon, 15 Mar 2004 19:39:47 +0100 Subject: [Stackless-sprint] first fprint summary version is online In-Reply-To: <4055DC62.5020302@gmx.net> References: <4055DC62.5020302@gmx.net> Message-ID: <4055F873.9090504@stackless.com> Lutz Paelike wrote: > A first version of a sprint summary is available: > > http://www.stackless.com/Members/lutz/SDocument.2004-03-15.0586225618/sdocument_view And if I try to read it, I get "insufficient privileges". This is a thing that I really hate to experience on my own website. I get less and less convinced that Plone is good for us. Let us come to a common decision on how to continue with the website, I feel a little uncomfortable right now. But this depends on what everybody else in the group thinks, so please give your opinion. cheers - chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From tismer at stackless.com Thu Mar 11 02:39:01 2004 From: tismer at stackless.com (Christian Tismer) Date: Thu, 11 Mar 2004 02:39:01 +0100 Subject: [Stackless-sprint] A real brain teaser [Fwd: Re: Stackless crash] Message-ID: <404FC335.1040505@stackless.com> Hi friends, trying to get this application to actually work, I already fixed a real bug with Stackless. But now things are no longer clear. It behaves differently in debug und release builds on Windows. Why and when does it hang after the sixth file, sometimes, and what's wrong in the shut down code whne we hit ctrl-c. Probably a mixture of problems, which also might be caused by recent changes to Python. see you all tomorrow -- chris p.s.: Tomek, didn't you say you would come to the sprint? -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ -------------- next part -------------- An embedded message was scrubbed... From: "Seung Chan Lim" Subject: Re: Stackless crash Date: Tue, 2 Mar 2004 00:56:52 -0500 Size: 20102 Url: http://www.stackless.com/pipermail/stackless-dev/attachments/20040311/f69a59d0/attachment.eml -------------- next part -------------- _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint From stackless-sprint-owner at stackless.com Tue Mar 2 18:12:44 2004 From: stackless-sprint-owner at stackless.com (stackless-sprint-owner at stackless.com) Date: Tue, 02 Mar 2004 18:12:44 +0100 Subject: Stackless-sprint post from senn@maya.com requires approval Message-ID: As list administrator, your authorization is requested for the following mailing list posting: List: Stackless-sprint at stackless.com From: senn at maya.com Subject: Re: [Stackless-sprint] Introduction Reason: Post by non-member to a members-only list At your convenience, visit: http://www.stackless.com/mailman/admindb/stackless-sprint to approve or deny the request. -------------- next part -------------- An embedded message was scrubbed... From: Jeff Senn Subject: Re: [Stackless-sprint] Introduction Date: Tue, 2 Mar 2004 12:09:18 -0500 Size: 1826 Url: http://www.stackless.com/pipermail/stackless-dev/attachments/20040302/5d0a428f/attachment.eml -------------- next part -------------- An embedded message was scrubbed... From: stackless-sprint-request at stackless.com Subject: confirm f37f3e0f4c7b11b7e820428c8116643ad6a36043 Date: no date Size: 648 Url: http://www.stackless.com/pipermail/stackless-dev/attachments/20040302/5d0a428f/attachment-0001.eml From tismer at stackless.com Thu Mar 18 17:54:00 2004 From: tismer at stackless.com (Christian Tismer) Date: Thu, 18 Mar 2004 17:54:00 +0100 Subject: [Stackless-dev] layout of stackless.com In-Reply-To: <31705E9C-78FA-11D8-A8B4-00039345C610@darwin.in-berlin.de> References: <31705E9C-78FA-11D8-A8B4-00039345C610@darwin.in-berlin.de> Message-ID: <4059D428.3040000@stackless.com> Dinu Gherman wrote: > Tomek Meka: > >> Well, Dinu just wanted to have glossary separetly from the >> whole wiki - that's why I've created the glossary folder. > > > Errmm, I wanted it to be a self-contained Wiki using Rest and > with the iterate script you wrote still working... ;-) Yes, I thought of more than one Wiki. One is the Wiki, one is the Glossary, and maybe more later? -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From tismer at stackless.com Tue Mar 16 16:49:32 2004 From: tismer at stackless.com (Christian Tismer) Date: Tue, 16 Mar 2004 16:49:32 +0100 Subject: [Stackless-dev] More thoughts on Greenlets Message-ID: <4057220C.4000801@stackless.com> Hi Armin, you might have seen my lengthy phantasies about greenlets. Btw.: I talked about "placable" greenlets, but I really meant "placeable" which is a huge difference :-) I haven't tried the module at all, but reading was very interesting. There is one place where I stumbled a little bit, and I guess it can be simplified: In g_initialstub you claim a certain piece of stack and jump off to the caller. The comment reads that you return twice. As stated elsewhere, I propose *not* to allocate the stack position on greenlet creation, but to do that when calling. I think, when a greenlet is created, it only saves the function to be called later and returns. When it is activated the first time, it is simply run in the current caller's stack and records its stack_start. Then, there are two possibilities: a) It does something, leaving state on the stack and jumps off to another greenlet. In this case it is bound to the piece of stack that it claims, and things work like you did it. b) It does something, stores all of its state in some structure, and then jumps to another greenlet without denoting its stack area. This means the greenlet is placeable again and can be run from anywhere without performing a jump. Does this create new problems? I see that you want to build this parent chain in the initialisation of the greenlet, to have a list which is ordered by stack address. But I think if this is done on the first activation, the condition of an ordered stack image still is fulfilled and we get much more flexibility. Pleae correct me if I'm wrong -- ciao - chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From stephan.diehl at gmx.net Thu Mar 18 10:48:53 2004 From: stephan.diehl at gmx.net (Stephan Diehl) Date: Thu, 18 Mar 2004 10:48:53 +0100 Subject: [Stackless-dev] layout of stackless.com Message-ID: <200403181048.53077.stephan.diehl@gmx.net> Dear friends, I probably should know, how everything is supposed to work on stackless.com, but I'm at a loss somehow. I have now a real unix account that works, so no problem there. Then I have the Plone account. There are still some security related problems, as I can't create new content or edit my own homepage. Can we use rest markup? Then there is still the Zope installation at www.centera.de. Is that still used? Is somebody moving at least the glossary over? (The Zope account is of course different to the Plone one). What kind of account is used with roundup? What's the layout of the CVS repository? If my plone access rights are fixed, I'll document your responses, of course. Cheers Stephan _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From gherman at darwin.in-berlin.de Thu Mar 18 17:34:57 2004 From: gherman at darwin.in-berlin.de (Dinu Gherman) Date: Thu, 18 Mar 2004 17:34:57 +0100 Subject: [Stackless-dev] layout of stackless.com In-Reply-To: <31600.1079626419@www61.gmx.net> Message-ID: <31705E9C-78FA-11D8-A8B4-00039345C610@darwin.in-berlin.de> Tomek Meka: > Well, Dinu just wanted to have glossary separetly from the > whole wiki - that's why I've created the glossary folder. Errmm, I wanted it to be a self-contained Wiki using Rest and with the iterate script you wrote still working... ;-) Dinu _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From tismer at stackless.com Fri Mar 26 17:21:34 2004 From: tismer at stackless.com (Christian Tismer) Date: Fri, 26 Mar 2004 17:21:34 +0100 Subject: [Stackless-dev] Re: weird behavior Re: [Stackless] New cvs version, new binaries In-Reply-To: <002101c412e3$978ccdc0$6401a8c0@djslim> References: <068a01c412d9$9fc55750$95fe46c0@GIRLS> <4063A240.8060600@stackless.com> <002101c412e3$978ccdc0$6401a8c0@djslim> Message-ID: <4064588E.8080306@stackless.com> Seung Chan Lim wrote: > gotcha, I didn't realize that stackless didn't intend to support threads. :) Well, I want to support them, but didn't do anything towards this, yet. It happens to work, but I'm not sure about shutdown conditions and such. > I'll try to get system threads out of the equation No problrm to get them in, again, but I need a concept, first. Design question for all: How should tasklets interact with threads? ciao - chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-dev mailing list Stackless-dev at stackless.com http://www.stackless.com/mailman/listinfo/stackless-dev From tismer at stackless.com Mon Mar 1 18:07:54 2004 From: tismer at stackless.com (Christian Tismer) Date: Mon, 01 Mar 2004 18:07:54 +0100 Subject: [Stackless-sprint] Introductions... In-Reply-To: References: Message-ID: <40436DEA.7030502@stackless.com> Richard Emslie wrote: ... > PS I need an address v. soon, so I can make travel arrangements etc.... > thanks. I was planning to take Bob and you into my house, if it is ok for you to share a room. We *could* take one more, but that would need sharing with a kid. Let's take that as a fallback solution. So the address in my .sig is your address. :-) Let me know about travel schedules, so I can pick you up. ciao - chris -- Christian Tismer :^) Mission Impossible 5oftware : Have a break! Take a ride on Python's Johannes-Niemeyer-Weg 9a : *Starship* http://starship.python.net/ 14109 Berlin : PGP key -> http://wwwkeys.pgp.net/ work +49 30 89 09 53 34 home +49 30 802 86 56 mobile +49 173 24 18 776 PGP 0x57F3BF04 9064 F4E1 D754 C2FF 1619 305B C09C 5A3B 57F3 BF04 whom do you want to sponsor today? http://www.stackless.com/ _______________________________________________ Stackless-sprint mailing list Stackless-sprint at stackless.com http://www.stackless.com/mailman/listinfo/stackless-sprint