- Pharo-dev - lists.pharo.org

Pharo 3 bugs on Linux with openWorldWithSpec?
by kmo Jan. 16, 2014

Jan. 16, 2014

I've come across two weird bugs (or at least strange effects) when using OpenWorldWithSpec. I'm not sure whether they are genuine bugs or it's just the way I've coded. I'm using the latest Pharo 3 on Ubuntu 13.10. I have a simple ComposableModel consisting of a TreeModel and a LabelModel which acts as a status line. The composableMorph is made up like this: spec <spec: #default> ^ SpecLayout composed add: #concordanceTreeModel origin: 0 @ 0 corner: 1 @ 1.0 offsetOrigin: 0 @ 0 offsetCorner: 0 @ self statusLineHeight negated; add: #statusLine origin: 0 @ 1.0 corner: 1.0 @ 1.0 offsetOrigin: 0 @ 0 offsetCorner: 0 @ self statusLineHeight negated; yourself When I run openInWorld I get the expected behaviour as below. <http://forum.world.st/file/n4735903/PharoScreenshot.2.png> Note how the background to the labelmodel status line is silver. But when I open the same ui with openWorldWithSpec I get this: <http://forum.world.st/file/n4735903/DarkGrayStatus.png> Note how the colour of the labelmodel background has changed. It's darker - in fact it has picked up the colour of my pharo desktop as though its background were transparent. Any ideas on this strange behaviour? Note also that the title of my ComposableModel is not copied to the pharo window either - this is bug 12231 which I reported a few weeks ago. The other strange effect is when I place the following code in a script |ui| ui := ConcordanceSpecUI new. ui openWorldWithSpec. and run it from the command line. This is what happens. <http://forum.world.st/file/n4735903/UsingStartupScript.png> (This is with an initialExtent specified - if there's no initial extent it's worse - I get a tiny area in the top left of the main window). But if I now click on the desktop, or resize the window, the interface suddenly corrects itself, like this: <http://forum.world.st/file/n4735903/good.png> I have tried this on Windows 7 and I do not get this behaviour - the ComposableModel fills the pharo window immediately and correctly. Any thoughts - should I be raising these as bugs? - Ken -- View this message in context: http://forum.world.st/Pharo-3-bugs-on-Linux-with-openWorldWithSpec-tp473590… Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

2 1

after an software update the current changeset is unnamed - literally
by Nicolai Hess Jan. 16, 2014

Jan. 16, 2014

This is what happens duing an update (30697 - 30698) save old cs Unnamed new changes 30698-Pha-Update new changes ScriptLoader30-MarcusDenker.853 new changes 30698-Pha-Update new changes Metacello-MC-MarcusDenker.696 new changes 30698-Pha-Update ChangeSet cleanup new changes Unnamed restore old cs <no name -- garbage?> new changes <no name -- garbage?> the old changeset is saved update stream starts-> every package update creates a new changeset scriptloader calls cleanup update stream ends old (cleanuped) cs is restored now every change goes to this unnamed changeset. If you open the changesorter it calls "ChangeSet current" which creates a new unnamed ("Unnamed") changeset and of course the prior changes are lost. Is this inevitable?

2 2

[pharo-project/pharo-core] d7228c: 30699
by GitHub Jan. 15, 2014

Jan. 15, 2014

Branch: refs/heads/3.0 Home: https://github.com/pharo-project/pharo-core Commit: d7228caeefd837d1bb66a541ed7e8171fdbe3561 https://github.com/pharo-project/pharo-core/commit/d7228caeefd837d1bb66a541… Author: Jenkins Build Server <board(a)pharo-project.org> Date: 2014-01-15 (Wed, 15 Jan 2014) Changed paths: M Morphic-Core.package/KeyboardEvent.class/instance/printing/storeOn_.st M Morphic-Core.package/WorldState.class/class/world menu items/windowsOn_.st M RPackage-Core.package/RPackageSet.class/instance/initialization/initialize_.st A ScriptLoader30.package/ScriptLoader.class/instance/pharo - scripts/script352.st A ScriptLoader30.package/ScriptLoader.class/instance/pharo - updates/update30699.st M ScriptLoader30.package/ScriptLoader.class/instance/public/commentForCurrentUpdate.st Log Message: ----------- 30699 12661 KeyboardEvent>>#storeOn: broken https://pharo.fogbugz.com/f/cases/12661 12660 MCPackage reviving RPackage when it shouldn't https://pharo.fogbugz.com/f/cases/12660 12659 Add Toggle full screen mode to World Windows menu https://pharo.fogbugz.com/f/cases/12659 http://files.pharo.org/image/30/30699.zip

1 0

[pharo-project/pharo-core]
by GitHub Jan. 15, 2014

Jan. 15, 2014

Branch: refs/tags/30699 Home: https://github.com/pharo-project/pharo-core

1 0

Re: [Pharo-dev] [squeak-dev] Constructing a patched Monticello package without installing it
by Eliot Miranda Jan. 15, 2014

Jan. 15, 2014

Thanks Colin! On Wed, Jan 15, 2014 at 12:06 PM, Colin Putney <colin(a)wiresong.com> wrote: > > > > On Wed, Jan 15, 2014 at 1:41 PM, Eliot Miranda <eliot.miranda(a)gmail.com>wrote: > > >> I just stumbled across Monticello's MCPatcher and the Backport button >> which claims to apply a partial delta to the ancestor of a package. Does >> anyone who understands this code know if the code supports loading a >> package, without installing it, applying a partial set of changes (i.e. the >> relevant changes form the bootstrap) and saving the package as an mcz? If >> it does then I could construct patched versions of the three packages and >> substitute them in the bootstrap instead of using the hack. >> > > Yeah, that ought to work fine. There's no way to do it through the GUI, > but a bit of workspace hackery should do the trick. > Here's a working example that patches a single method: | fn fs version method patch ancestry newVersion | fn := '/Users/eliot/Glue/repositories/spurnsboot/smalltalk/Kernel.newspeak-bwesterg.682.mcz'. fs := FileStream readOnlyFileNamed: fn. version := [((MCVersionReader readerClassForFileNamed: fn) on: fs fileName: fn) version] ensure: [fs close]. method := SpurBootstrap class >> #ProtoObjectPROTOTYPEscaledIdentityHash. patch := MCPatch operations: {MCAddition of: (MCMethodDefinition className: #ProtoObject classIsMeta: false selector: #scaledIdentityHash category: #comparing timeStamp: method timeStamp source: (method getSourceFromFile asString allButFirst: 20))}. ancestry := MCWorkingAncestry new addAncestor: version info. newVersion := MCVersion package: version package info: (ancestry infoWithName: 'Kernel.newspeak.spur-bwesterg.682.mcz' message: 'Kernel.newspeak-bwesterg.682.mcz patched for Spur') snapshot: (MCPatcher apply: patch to: version snapshot) dependencies: {}. (MCDirectoryRepository new directory: fs directory) storeVersion: newVersion -- best, Eliot

1 0

[OT] Interviews of RMoD team members
by Nicolas Petton Jan. 15, 2014

Jan. 15, 2014

Hi guys, Sorry for this off topic post, but I thought it might be interesting for people following Pharo's development :) I started a series of interviews of RMoD team members. The first one is the interview of Damien Cassou, online here http://nicolas-petton.fr/blog/interview-damien.html Cheers, Nico

2 2

Spoon newsgroup
by Craig Latta Jan. 15, 2014

Jan. 15, 2014

Hi all-- I've posted links to some recent Spoon posts to the Spoon newsgroup, gmane.comp.lang.smalltalk.squeak.spoon, from my blog, thiscontext.com. I'd love to converse there. thanks! -C -- Craig Latta www.netjam.org/resume +1 510 984 8117 (Skype rings this until 31 January 2014)

1 0

Constructing a patched Monticello package without installing it
by Eliot Miranda Jan. 15, 2014

Jan. 15, 2014

Hi All, I'm trying to deploy Spur in our Newspeak bootstrap at Cadence. The starting point for the Newspeak bootstrap) is a Squeak4.3 image. I have a Squeak4.3 image transformed by the Spur bootstrap to Spur format, which means it has a few method modifications to three packages: Collections (changed Character methods now that Characters are immediate) Kernel (changed Behavior & ClassBuilder changes now that a class's format has changed, and instantiation interacts with the garbage collector differently) System (changed SystemDictionary/SmalltalkImage methods now that there's a generation scavenger) The Newspeak bootstrap loads versions of these packages that support Newspeak. So the problem is to load the Newspeak versions of these packages without undoing the Spur bootstrap modifications. As a quick hack I could patch the Monticello loader to avoid the modifications (they're marked by not having a source pointer but could be marked in some other way, e.g. a pragma). That's a fine simple approach. But... I just stumbled across Monticello's MCPatcher and the Backport button which claims to apply a partial delta to the ancestor of a package. Does anyone who understands this code know if the code supports loading a package, without installing it, applying a partial set of changes (i.e. the relevant changes form the bootstrap) and saving the package as an mcz? If it does then I could construct patched versions of the three packages and substitute them in the bootstrap instead of using the hack. -- best, Eliot

1 0

Beware Issue 12657 Nautilus: Method DND explodes
by Sean P. DeNigris Jan. 15, 2014

Jan. 15, 2014

Nasty one: Drag method from one open Nautilus to a class in another browser. 1. The method disappears from the original class, but is not listed as removed in MC changes 2. The method does not appear in the target class ----- Cheers, Sean -- View this message in context: http://forum.world.st/Beware-Issue-12657-Nautilus-Method-DND-explodes-tp473… Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

1 0

Recordable bug reports
by Sean P. DeNigris Jan. 15, 2014

Jan. 15, 2014

Kind of a cool use for EventRecorderMorphâ¦ I noticed this method comment: "In case there is a chance of not regaining control to stop recording and save a file, the EventRecorder can write directly to file as it is recording. This is useful for capturing a sequence that results in a nasty crash." You could record the steps to reproduce a bug with EventRecorderMorph, save it to a tape file, and attach it to the bug report. ----- Cheers, Sean -- View this message in context: http://forum.world.st/Recordable-bug-reports-tp4736970.html Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

2 1