January 2014 - Pharo-dev - lists.pharo.org

some hand holding with git and the pharovm tree please...
by Eliot Miranda Jan. 24, 2014

Jan. 24, 2014

Hi All, in trying to help ClÃ©ment with his Sista work on Cog I'm trying to look at the latest version of pharovm. I've been using git:// gitorious.org/cogvm/blessed. I cloned it on December 2nd, and just did a git fetch, but I see no changes to any files other than .git/FETCH_HEAD and .git/ORIG_HEAD. How do I update the files to their current state (I use svn & mercurial, have used sccs, rcs & sccs, and it's such a pain having to learn yet another f***ing sccs)? Further, I see no generated files such as gcc3x-cointerp.c or cogit.c. Why not? Do you always fire up an image and freshly generate the C source?? If not, where do I find it? All this is to answer the simple question of whether the blessed pharo vm contains primitiveInterpreterSourceVersion. How would you answer that question without running the VM? e.g. in the Cog svn repository you'd do grep primitiveInterpreterSourceVersion src/vm/*.c src/plugins/*/*.c and you'd have an answer. I tried cd mc/VMMaker-oscog.package/instance grep primitiveInterpreterSourceVersion *.st and the lack of reply makes me think it doesn't exist. Hence from !VirtualMachine methodsFor: 'accessing' stamp: 'MarcusDenker 5/8/2013 11:51'! interpreterSourceVersion "The use of this primitive not recommended. Not all VMs providing that" "Answer a string corresponding to the version of the interpreter source. This represents the version level of the Smalltalk source code (interpreter and various plugins) that is translated to C by a CCodeGenerator, as distinct from the external platform source code, typically written in C and managed separately for each platform. An optional primitive is invoked that may not be available on all virtual machines." "Smalltalk vm interpreterSourceVersion" <primitive: 'primitiveInterpreterSourceVersion'> ^ self getSystemAttribute: 1009! ! I can tell I need to look at the implementation of sourceVersionString in platforms/Cross/vm/sqSCCSVersion.h, which, despite all my best efforts, has diverged completely (and IMO unnecessarily) from http://www.squeakvm.org/svn/squeak/branches/Cog/platforms/Cross/vm/sqSCCSVe…, (which was written to be extended to satisfy us all). It reads /* * A set of definitions for C source code control systems, to provide accurate * and definitive version information to the VM. * * Currently instantiated only for Subversion. Please add definitions for * other repositories as appropriate. * * I guess a good way to manage this is to edit the below define list to select * appropriate the repository type, and then that's the extent of the fork. * * Eliot Miranda * eliot.miranda(a)gmail.com * 15 July 2011 */ #include "vmVersionInfo.h" // defines REVISION_STRING #ifndef REVISION_STRING # error "Revision information not specified" #endif static char *sourceVersionString() { return REVISION_STRING; } which is wrong. I *didn't* write this. I wrote something quite different and you should have edited it to reflect that. But alas, when I try find . -iname vmVersionInfo.h I find...nothing. hmph. stymied. frustrated. confused. hence this missive. P.S> What I'm really trying to do is get my VM simulator to say it's a Pharo VM, so I can simulate ClÃ©ment's Sista image and fix the infrastructure he's trying to use for speculative inlining/adaptive optimization. So this is in aid of a (much) faster Pharo VM. Hence helpful answers only please. -- best, Eliot

3 2

Nautilus Package Tree
by Sean P. DeNigris Jan. 24, 2014

Jan. 24, 2014

Is more-than-one-level deep planned? I love the idea, but I'm not finding the current implementation particularly useful. I want to reiterate that I think the tree nodes should be based on name-matching and not only per-package. For example: > MyProject >X Core - Support - Announcements X Tests X Platform The icons (simulated by X's above) serve to distinguish what is a package and what is not. Especially for new users who will be overwhelmed, at the top level of the system, when one is not working on MyProject, seeing a single MyProject node is the right thing: #1: > Keymapping > MyProject > NativeBoost instead of the current (even when collapsed): #2: > Keymapping-Core Keymapping-KeyCombinations Keymapping-Pragmas Keymapping-Settings Keymapping-Tests Keymapping-Tools-Spec > MyProject-Core > MyProject-Tests MyProject-Platform > NativeBoost-Core > NativeBoost-Examples > NativeBoost-Mac > NativeBoost-Pools > NativeBoost-Tests > NativeBoost-Unix > NativeBoost-Win32 #1 captures the feel of the system from a birds eye view, but even with only 3 high-level concept spaces, #2 is already double human RAM. The fact that there is a MyProject-Tests package, which is only separate so that it can be unload in production, and that there is a MyProject-Platform that needs to be a separate package so that we can support both Squeak and Pharo, is irrelevant and distracting from that point of view. With the current tree-root per package, there are over 250 top-level nodes. This has not made understanding and navigating the system any more sane. Maybe the use of pattern matching as make-believe packages for so long has made us wary of the technique, but let's use each tool where it's appropriate. ----- Cheers, Sean -- View this message in context: http://forum.world.st/Nautilus-Package-Tree-tp4738887.html Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

1 1

RFB on a headless Pharo 3
by Sebastian Sastre Jan. 23, 2014

Jan. 23, 2014

Guys Just checking while investigating the issue... Were any of you here able to connect a VNC client to a headless Pharo 3? I can do that to a Pharo 2 (seting a morphic ui manager on it) but I'm having issues with 3 Did you tried that? any luck? insights? sebastian o/

2 7

Senders in Scoped Browser Aren't
by Sean P. DeNigris Jan. 23, 2014

Jan. 23, 2014

1. Select a few packages in Nautilus 2. Browse scoped from the context menu 3. in the class definition pane, type a selector (I chose #new) 4. Browse senders via the shortcut They are not scoped ----- Cheers, Sean -- View this message in context: http://forum.world.st/Senders-in-Scoped-Browser-Aren-t-tp4738789.html Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

1 0

Services and -vm-display-null
by Sebastian Sastre Jan. 23, 2014

Jan. 23, 2014

Hi there, What are you using to monitor processes in headless images? I ask because the new way to start a headless image in linux is: pharo-vm/pharo -vm-display-null your.image âno-quit And it seems that it always quits For example supervisord* is really unhappy about this way to run a service. Itâs really unfortunate. This is a show stopper. Is there an OS signal being sent when you use -vm-display-null or the process stays normally up? There is no way to run a headless image like we used to? sebastian o/ *http://supervisord.org/

4 5

Versioning - Monticello vs. Bitbucket/Github/Gitorious etc "how to" lesson at ESUG14?
by Sabine KnÃ¶fel Jan. 23, 2014

Jan. 23, 2014

I just read this: http://smallworks.com.ar/web/blog/2014-01-16-Why-Im-using-git Very interesting, thank you, Esteban. Wouldn't it be good to have a lesson about this topic at ESUG14? Sabine -- View this message in context: http://forum.world.st/Versioning-Monticello-vs-Bitbucket-Github-Gitorious-e… Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

1 0

PSMCPatchMorph - can I remove the setting?
by Sean P. DeNigris Jan. 23, 2014

Jan. 23, 2014

What is "PSMCPatchMorph usedByDefault" all about? When would we not want to use the polymorph diff tool for MC? If never, I'd like to remove both the setting and MCMergeBrowser ----- Cheers, Sean -- View this message in context: http://forum.world.st/PSMCPatchMorph-can-I-remove-the-setting-tp4738759.html Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

2 3

Forward IRC Channel
by Sean P. DeNigris Jan. 23, 2014

Jan. 23, 2014

Whoever has access, maybe it would be good to: /cs redirect #pharo-project #pharo [1] As suggested to me by eMBee, the only other user in #pharo-project ;) [1] http://www.geekshed.net/2010/05/redirecting-one-channel-to-another/ ----- Cheers, Sean -- View this message in context: http://forum.world.st/Forward-IRC-Channel-tp4738764.html Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

1 0

Morphic Alarms
by Sean P. DeNigris Jan. 23, 2014

Jan. 23, 2014

Is anyone using them? What is a real world use case? I've read through the class comment, senders (currently seems to be only DebuggerSession), etc... ----- Cheers, Sean -- View this message in context: http://forum.world.st/Morphic-Alarms-tp4738603.html Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

2 1

#windowIsClosing not sent to model of WindowModel
by btc＠openinworld.com Jan. 23, 2014

Jan. 23, 2014

PharoLauncher is currently leaking AnnouncementSubscriptions when its window closes. I'm not sure with Spec how to ensure actions are performed when the window is closed, but I have been able to achieve this with the following modification [1] which is... ----- WindowModel>>windowIsClosing isClosedHolder value: true self model windowIsClosing "<----proposed modification" ----- I investigated with a halt in SpecWindow>>close to trace through and discover that it calls... SpecWindow(StandardWindow)>>delete, which in my non-fullscreen-case calls... SpecWindow(SystemWindow)>>delete, which "model windowIsClosing" calls... MorphicWindowAdaptor>>windowIsClosing, which "self model windowIsClosing" calls... WindowModel>>windowIsClosing, which does only "isClosedHolder value: true" That last line happens to hold aPharoLauncher in its 'model' instance variable, so with the above proposal I can implement PharoLauncher>>windowIsClosing to unregister the announcer. Any problem expected with that modification? [1] https://pharo.fogbugz.com/f/cases/12677 cheers -ben

2 10