February 2012 - Pharo-dev - lists.pharo.org

Re: [Pharo-project] garbage collection

by Sven Van Caekenberghe

Milan, On 14 Feb 2012, at 21:01, Milan Mimica wrote: > This is more of a GC theory question perhaps, but I do have practical problems, so: I'm trying to figure out how come my headless Pharo 1.3 image with fairly recent Cog running Magma ran out of Semaphores. I know there is a fixed limit on number of semaphores Cog can handle, and only Magma and RFB server have been running on this image for two months, and there wasn't any heavy load whatsoever, so something must be leaking semaphores - or more likely - something must be leaking objects having reference(s) to semaphore(s). I am not leaking sockets btw. > > At the moment there is about 1400 instances of Semaphore class, and around 400 after GC. > So first question is: Suppose I haven't run GC manually, could the semaphore limit be hit? > > 400 is still too high for a idling image, don't you think? > > Anyway, what I try next is to shut down Magma completely and try to clean the memory by hand. I notice some MaServerSockets are still floating around so I track the pointers and I find that all the references are circular! > Question two: Does Pharo's GC solve circular references? If yes, how good is it at it? > > But then, thinking more about it, the whole running image is one big graph of circular references, you wouldn't want the GC to clean that :) So how does it really work? That's my naive view anyway. > > After I wrote all this I figure I'm better of taking a fresh new image and load what I need into it, even if I have to do it every two months. Not saving the image is also an option. Yeah, I think I'm going to do just that. I found my solution, at least this e-mail served for something. Since I don't have a blog I'll send it anyway. > > > -- > Milan Mimica There was a semaphore related problem in 1.3 that was fixed some time ago. Doing regular image saves was one of the factors involved. Search the mailing list or issues list for semaphore. Sven

Feb. 14, 2012

[Pharo-project] garbage collection

by Milan Mimica

This is more of a GC theory question perhaps, but I do have practical problems, so: I'm trying to figure out how come my headless Pharo 1.3 image with fairly recent Cog running Magma ran out of Semaphores. I know there is a fixed limit on number of semaphores Cog can handle, and only Magma and RFB server have been running on this image for two months, and there wasn't any heavy load whatsoever, so something must be leaking semaphores - or more likely - something must be leaking objects having reference(s) to semaphore(s). I am not leaking sockets btw. At the moment there is about 1400 instances of Semaphore class, and around 400 after GC. So first question is: Suppose I haven't run GC manually, could the semaphore limit be hit? 400 is still too high for a idling image, don't you think? Anyway, what I try next is to shut down Magma completely and try to clean the memory by hand. I notice some MaServerSockets are still floating around so I track the pointers and I find that all the references are circular! Question two: Does Pharo's GC solve circular references? If yes, how good is it at it? But then, thinking more about it, the whole running image is one big graph of circular references, you wouldn't want the GC to clean that :) So how does it really work? That's my naive view anyway. After I wrote all this I figure I'm better of taking a fresh new image and load what I need into it, even if I have to do it every two months. Not saving the image is also an option. Yeah, I think I'm going to do just that. I found my solution, at least this e-mail served for something. Since I don't have a blog I'll send it anyway. -- Milan Mimica

Feb. 14, 2012

Re: [Pharo-project] Feedback from lecture :(

by Igor Stasenko

On 14 February 2012 20:22, Lukas Renggli <renggli(a)gmail.com> wrote: >>> students in the lecture on windows downloaded cogvm I do not from whereâ¦ >>> and used pharo1.3 and >>> some experienced crashes. >> >> This has been reported at least 10 times per week since the last 6 months. >> The VMs of the one click are FUC.. old and buggy. If nobody have time to >> make a new one click (we are all busy), please please please, remove it from >> the website. We give so much bad impression. It is worst than having no one >> click. > > You cannot say that, because I updated the One-Click distribution less > than a month ago: > > Â https://github.com/renggli/builder/commit/7623726096db2e3fa58eefe267c62ba17… > > People using this code get the latest available CogVM. The problem is > that Pharo maintains its own unmaintained fork. > Huh? See the build history: https://ci.lille.inria.fr/pharo/view/Cog/job/Cog-Unix/ last build 26 Jan https://ci.lille.inria.fr/pharo/view/Cog/job/StackVM-Mac-Cocoa-blessed/ again, last build 28 Jan. Here just a list of changes in latest builds: (you can find them here: https://ci.lille.inria.fr/pharo/view/Cog/job/Cog%20Git%20Tracker%20%28bless… ) Changes #26 (Jan 26, 2012 5:33:21 PM) replaced spaces with tabs â camillobruni / detail #25 (Jan 26, 2012 5:18:48 PM) added new pharo scropt â camillobruni / detail - fixed configuraiton to open a Workspace with example configurations â camillobruni / detail trying to fix issues with the command line arguments â camillobruni / detail more work on the image script and the howto â camillobruni / detail tried to implemention dir_lookupEntry for Cococa OSX â camillobruni / detail extended howto with osx gcc change â camillobruni / detail fixed back opengl changes â camillobruni / detail fixed VM path, returning bundlepath with app name â camillobruni / detail implemented direEntryLookup and refactored a bit the source code â camillobruni / detail more howto â camillobruni / detail minor change in the HOWTO â camillobruni / detail undoing accidental changes â camillobruni / detail #23 (Jan 25, 2012 8:06:27 PM) removed deprecated calls to recordKeystroke â guillermopolito / detail removed unused comment â guillermopolito / detail removed old code â guillermopolito / detail removed dead code â guillermopolito / detail removed dead code â guillermopolito / detail removed lot of debug only code â guillermopolito / detail removed temporal file â guillermopolito / detail removed more debug printfs â guillermopolito / detail formatted the code â guillermopolito / detail little style change â guillermopolito / detail added mappings to numeric keypad â guillermopolito / detail added comments to unicode conversion â guillermopolito / detail more comments â guillermopolito / detail For the full list of changes in VM , visit https://gitorious.org/cogvm/blessed/commits/master but just don't say bullshit like unmaintained unmaintained VM. Now if people do not downloading it, do not report problems, and sticking to Eliot's VMs, it makes no sense to do anything. You know well, that we cannot force people to use it or not. (i triggered new build now just to check if everything ok). -- Best regards, Igor Stasenko.

Feb. 14, 2012

[Pharo-project] Who can help for the pharo web site

by stephane ducasse

Hi guys Adrian left and now I cannot manage the pharo web site. I need help because I do not understand how to edit it. Today by accident I removed the (broken) twitter box. So what do we do? I was thinking to port everything in plain static html so that I can edit the pages fast and easy. I would like to add a how to contribute page. Our web site does not evolve anymore because nobody cares about it. So I can try to do it but I will port everything into html. Stef (tired to do too many things and fed up about systematic bashing).

Feb. 14, 2012

Re: [Pharo-project] Are Objects really hard?

by Andreas Wacknitz

Am 14.02.2012 um 20:01 schrieb Guido Stepken: > Get the development process developed! > Guido, you seem to have interest in Pharo, even if it doesn't fit your needs today. You also seem to have "the right" connections and money. What about turning this into something positive for you, your customers and the whole community? Make a list of things you want to have. Make the things more concrete than in your last posts. An wave with a bundle of money. You will maybe raise the interest of some of the brilliant guys here and can convince them to do what you want. From what you wrote within this thread I would bet that you will be able to offer more than $100,000 - a very convincing argument and maybe enough to get what you want from Pharo! Regards, Andreas

Feb. 14, 2012

Re: [Pharo-project] Feedback from lecture :(

by Guido Stepken

Am 14.02.2012 um 20:10 schrieb Igor Stasenko > What is LLVM? > Here , read: > > The LLVM Project is a collection of modular and reusable compiler and > toolchain technologies. Despite its name, LLVM has little to do with > traditional virtual machines, though it does provide helpful libraries > that can be used to build them. Read: http://fosdem.org/2012/interview/david-chisnall tnx 4 understanding, Guido Stepken

Feb. 14, 2012

Re: [Pharo-project] Feedback from lecture :(

by StÃ©phane Ducasse

On Feb 14, 2012, at 8:22 PM, Lukas Renggli wrote: >>> students in the lecture on windows downloaded cogvm I do not from whereâ¦ >>> and used pharo1.3 and >>> some experienced crashes. >> >> This has been reported at least 10 times per week since the last 6 months. >> The VMs of the one click are FUC.. old and buggy. If nobody have time to >> make a new one click (we are all busy), please please please, remove it from >> the website. We give so much bad impression. It is worst than having no one >> click. > > You cannot say that, because I updated the One-Click distribution less > than a month ago: > > https://github.com/renggli/builder/commit/7623726096db2e3fa58eefe267c62ba17… > > People using this code get the latest available CogVM. The problem is > that Pharo maintains its own unmaintained fork. It is not unmaintained. Now if really few people help then igor cannot do everything. And since eliot does not care about the effort we spent to make sure that people can compile their own vms then we have to life with that. Also the pharo vms has important fixes for us. > Lukas > > -- > Lukas Renggli > www.lukas-renggli.ch >

Feb. 14, 2012

Re: [Pharo-project] Feedback from lecture :(

by Igor Stasenko

On 14 February 2012 20:04, Mariano Martinez Peck <marianopeck(a)gmail.com> wrote: > > > On Tue, Feb 14, 2012 at 7:35 PM, StÃ©phane Ducasse > <stephane.ducasse(a)inria.fr> wrote: >> >> students in the lecture on windows downloaded cogvm I do not from whereâ¦ >> and used pharo1.3 and >> some experienced crashes. > > > This has been reported at least 10 times per week since the last 6 months. > The VMs of the one click are FUC.. old and buggy. If nobody have time to > make a new one click (we are all busy), please please please, remove it from > the website. We give so much bad impression. It is worst than having no one > click. > It was known before we released one-click, that VM is not perfect. But we cannot delay the release because of that. It would make things only worse. Because every time you going to release, you can find one or another issue which needs to be fixed before release. What is important to understand, that we don't releasing a final product, the Pharo releases are milestones on continuing development. Yes, every release having issues and they cannot be perfect ,and there are always will be things which left not done/not fixed etc etc. So we need to accept that. >> >> >> We will have to brand these vms because the situation is getting >> difficult. > > > No, that's a second step. Put it in Jenkins is the third. Do the first step: > do a new OneClick BY HAND, with latest Cog, even without "pharo blessing". > That way new comers will not run away from pharo beacuse it crashes. > And why you think that new one will be more stable that old? i just recently built a windows VM. it crashes occasionally/randomly in some bitblt primitive(s) during drawing a progress bar when installing MC package, i have no idea why. I suspect a memory corruption somewhere when new classes installed in a system , or existing classes are modified by MC, and bitblt prim is just a consequence of it. Which not saying that we should not address this issue(s), but that latest version is not necessary more stable than older one. To guarantee that there are no regressions like that, we need to setup good testing infrastructure around VM. And while we're still not there.. but will be soon. > > >> >> I do not imagine how we can track these problems. >> >> Stef > > > > > -- > Mariano > http://marianopeck.wordpress.com > -- Best regards, Igor Stasenko.

Feb. 14, 2012

Re: [Pharo-project] Feedback from lecture :(

by StÃ©phane Ducasse

> On Tue, Feb 14, 2012 at 7:35 PM, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: > students in the lecture on windows downloaded cogvm I do not from whereâ¦ and used pharo1.3 and > some experienced crashes. > > This has been reported at least 10 times per week since the last 6 months. > The VMs of the one click are FUC.. old and buggy. If nobody have time to make a new one click (we are all busy), please please please, remove it from the website. We give so much bad impression. It is worst than having no one click. > > > We will have to brand these vms because the situation is getting difficult. > > No, that's a second step. Put it in Jenkins is the third. Do the first step: do a new OneClick BY HAND, with latest Cog, even without "pharo blessing". That way new comers will not run away from pharo beacuse it crashes. Where can I find the correct VMs? I can give a try. Now I did not do it before and I'm really busy. > > > > I do not imagine how we can track these problems. > > Stef > > > > -- > Mariano > http://marianopeck.wordpress.com >

Feb. 14, 2012

Re: [Pharo-project] Nautilus Refactorings

by Benjamin

Basically, the infrastructure is here. But I still have to copy and past each menu entry :) It's long and a bit boring, but it's easy to do ^^ Ben On Feb 14, 2012, at 8:15 PM, Noury Bouraqadi wrote: > Cool. I'll give it a try. I'm like mariano, luc and other waiting for RB support to start using Nautilus as the default browser. > > Noury > On 14 fÃ©vr. 2012, at 18:39, Benjamin wrote: > >> Hello guys, >> >> I know a some of you are waiting for the integration of the RB engine into Nautilus. >> >> Some of them are already implemented, so if you wanna test them, here is the gofer script >> >> Gofer new >> url: 'http://ss3.gemstone.com/ss/Nautilus'; >> package: 'ConfigurationOfNautilus'; >> load. >> >> (Smalltalk at: #ConfigurationOfNautilusRefactoring) perform: #loadDefault >> >> >> It loads everything. >> >> For now I have only done some refactorings for classes, so if you popup the menu on a class, a new item named Refactoring should appear :) >> >> Moreover, the 'rename class' method is replaced by the RB one :) >> >> >> Thanks in advance for your feedback, and if you want to participate, you are more than welcome :) >> >> >> Ben > > Noury > -- > http://twitter.com/#!/NouryBouraqadi > http://www.kroobe.com/profile/noury > >

Feb. 14, 2012