November 2017 - Pharo-dev - lists.pharo.org

VoxelSpace
by Pavel Krivanek Nov. 29, 2017

Nov. 29, 2017

Hi, I tried to port and optimize the extremely simple VoxelSpace engine into Pharo. https://github.com/pavel-krivanek/VoxelSpace How to load Metacello new baseline: 'VoxelSpace'; repository: 'github://pavel-krivanek/VoxelSpace/src'; load. How to run Copy maps into the image directory and then evaluate: VoxelSpaceMorph new openInWindow. [image: VloÅ¾enÃ½ obrÃ¡zek 2] No external library is required. If you have some next ideas how to improve the rendering performace with the same image quality, please let me know. Cheers, -- Pavel

3 4

(no subject)
by Ben Coman Nov. 29, 2017

Nov. 29, 2017

A minor thing, but a fix is probably easy... I have... FFIExternalStructure subclass: #CXSourceLocation instanceVariableNames: '' classVariableNames: 'VoidPointer2' package: 'Libclang' initialize "self initialize" VoidPointer2 := FFITypeArray ofType: 'void*' size: 2. fieldsDesc "self rebuildFieldAccessors" ^#( VoidPointer2 ptr_data; uint int_data; ) Now when I debug into "self rebuildFieldAccessors" into #compileFields:withAccessors: through "fieldSpec fieldsAndTypesDo: [ :fieldName :type | " the [Variables] tab shows... type "FFITypeArrayType(a subclass of FFITypeArray)" and my perception is that this is wrong since FFITypeArrayType is not a subclass FFITypeArray. Intuitively I'm reading the text in the round brackets as a commentary on FFITypeArrayType. But debugging into "type printString" bring me to FFITypeArrayType(FFIExternalReferenceType) >> printOn: aStream aStream << self className << '(' << objectClass name << ')' where objectClass is an instance variable, so the commentary is about composition rather than inheritance. Now I properly understand the truth, I still have trouble with perceiving the printString the wrong way, so I'd like to add a visual cue that this is composition. Do we have such a standard convention? Options... 1. Square brackets a. FFITypeArrayType[a subclass of FFITypeArray] b. FFITypeArrayType([a subclass of FFITypeArray]) 2. Curly brackets a. FFITypeArrayType{a subclass of FFITypeArray} b. FFITypeArrayType({a subclass of FFITypeArray}) 3. Array-ish a. FFITypeArrayType#(a subclass of FFITypeArray) b. FFITypeArrayType(#(a subclass of FFITypeArray)) 4. Text a. FFITypeArrayType(holding a subclass of FFITypeArray) I'm tending towards [2a] or [2b] since two brackets together seem more distinct. It may be easy to overlook a difference in a single bracket, and also it feels more like holding something inside the object. cheers -ben P.S. Regarding the existing Libclang bindings. I'm not trying to compete but more along the line... "its worthwhile reinventing the wheel when you want to learn more about how a wheel works." I'll switch over to existing bindings in a while.

1 0

[Pharo 7.0-dev] Build #346: 20777-Fix-instVarAt-instVarAt-put-comment
by ci-pharo-ci-jenkins2＠inria.fr Nov. 28, 2017

Nov. 28, 2017

There is a new Pharo build available! The status of the build #346 was: FAILURE. The Pull Request #559 was integrated: "20777-Fix-instVarAt-instVarAt-put-comment" Pull request url: https://github.com/pharo-project/pharo/pull/559 Issue Url: https://pharo.fogbugz.com/f/cases/20777 Build Url: https://ci.inria.fr/pharo-ci-jenkins2/job/Test%20pending%20pull%20request%2…

1 0

FFIIntegerTypes inheritance hierarchy
by Ben Coman Nov. 28, 2017

Nov. 28, 2017

I'm looking through the keyhole at something new to me, so maybe this is off base, but looking at the inheritance hierarchy... FFIIntegerType FFIUint16 FFIUint16 FFIUInt32 FFIInt32 FFIUint64 FFIInt64 FFIUint8 FFIInt8 I'm curious... why the all the Ints inherit from the UInts? Each of those classes has the same two methods #basicHandle:at: and #basicHandle:at:put. So the Int subclasses fully override the superclass methods. So there in no inherited behaviour on the instance side. On the class side, I see #externalTypeSize is inherited by the Ints, but evaluating... FFIIntegerType allSubclasses collect: [ :class | class externalType byteSize -> class externalTypeSize ] produces... 4->4 2->2 4->4 8->8 1->1 2->2 4->4 8->8 1->1 So naively the extra inheritance level seems unnecessary and perhaps having... FFIIntegerType class >> externalTypeSize ^ self externalType byteSize would allow the hierarchy to be flattened. Maybe not a great benefit, except to avoid the next person wondering the same thing :P Also, it is surprising that that FFUInt64 uses #signedLongLongAt: rather than #unsignedLongLongAt: . This seems on-purpose rather than accidental since its inherited by FFIInt64, so it seems there is something interesting to learn here that would benefit a method comment. cheers -ben

2 2

Would it be possible to also launch templates in PharoLauncher?
by Mariano Martinez Peck Nov. 28, 2017

Nov. 28, 2017

Hi guys, Thank you Christophe for continue pushing this useful tool! One small request which might be easy to do... Quite frequently I want to do something and re-save an image I have previosuly saved as a custom template (under Local root tree). Right now what I have to do so: 1) Make a new image from that template 2) remove template 2) do the thing and save the opened image. 3) go back to laucher and save that image as a new template (with the same name of the one I deleted) 4) delete the created image All that would be simplified if I could also launch my local templates. Is there any reason why is that not allowed? (at least for Local tree). Thanks in advance, -- Mariano http://marianopeck.wordpress.com

3 4

[Pharo 7.0-dev] Build #345: 20777-Fix-instVarAt-instVarAt-put-comment
by ci-pharo-ci-jenkins2＠inria.fr Nov. 28, 2017

Nov. 28, 2017

There is a new Pharo build available! The status of the build #345 was: FAILURE. The Pull Request #559 was integrated: "20777-Fix-instVarAt-instVarAt-put-comment" Pull request url: https://github.com/pharo-project/pharo/pull/559 Issue Url: https://pharo.fogbugz.com/f/cases/20777 Build Url: https://ci.inria.fr/pharo-ci-jenkins2/job/Test%20pending%20pull%20request%2…

1 0

Iceberg "Create repository" buttons
by Ben Coman Nov. 28, 2017

Nov. 28, 2017

In Iceberg I am finding the "Create repository" button disconcerting for some of the Iceberg repository dialogs. For example, clicking "Add local repository" shows a dialog "Import local repository into Iceberg" with a "Create repository" button and perception of this which I can't shake is "Whoa! I don't want to create a repository. It already exists!" Perhaps this is a narrow view, but for me a git repository exists properly on a remote or under the ".git" folder on the file system. Repositories don't exist inside the Image, only a single checked out working version. So "Create repository" feels like it can only operating on the file system. This is a left-brain/right-brain thing. Technically I can reason its the only button so pushing it can't be wrong, but it "feels" like pushing it will do the wrong thing. wdyt? cheers -ben

1 0

[Pharo 7.0-dev] Build #344: 20772 - Unused temps in MBConfigurationInfo, MethodDictionaryTest, ...
by ci-pharo-ci-jenkins2＠inria.fr Nov. 28, 2017

Nov. 28, 2017

There is a new Pharo build available! The status of the build #344 was: SUCCESS. The Pull Request #557 was integrated: "20772 - Unused temps in MBConfigurationInfo, MethodDictionaryTest,..." Pull request url: https://github.com/pharo-project/pharo/pull/557 Issue Url: https://pharo.fogbugz.com/f/cases/20772 Build Url: https://ci.inria.fr/pharo-ci-jenkins2/job/Test%20pending%20pull%20request%2…

1 0

QA rule suggestion - assignment symbol slips
by Ben Coman Nov. 28, 2017

Nov. 28, 2017

A quick suggestion for anyone working on QA rules... Just now I am playing with FFI a bit and switching back a forth from Pharo to C and sometimes I end up doing... location = Libclang clang_getRangeStart__range: range. thus QA reported a. "Temporaries read before written" and b. "Temporary variables not read or not written" but these weren't helpful since I was stuck staring at it saying to myself "Well range is assigned before its used, and location is assigned, so bah! the rules seem broken" and for about 30 seconds I remained blind to the truth I'd left the colon off the assignment operator. I'd expect newcomers from other languages might slip in the same way, so a higher priority rule extending [b.] that identifies such variable followed by equal-sign would provide pertinent feedback. cheers -ben

2 4

clarifying pharo crashing issue(s)
by Tudor Girba Nov. 28, 2017

Nov. 28, 2017

Hi, There are recurrent reports that Pharo is crashing often. However, at the moment I believe the reasons are not clarified. So, this mail is not about fixing the issues, but first about clarifying what they are. Namely: - I know about the typing problem on MacOS (Sierra or High Sierra). I encounter this problem daily (literally), and I know that I am not the only one. For example, this can occur when I type âclasssâ fast and then delete the last character. It is documented here: https://github.com/OpenSmalltalk/opensmalltalk-vm/issues/141 - Some people talk about the FreeType problem, but I could not find a place where this is described. So, while this might be a problem, I am not sure that this problem is actually causing crashes. So, first question: - is Pharo crashing on another system than MacOS Sierra or HighSierra? - how does the FreeType problem look like? Cheers, Doru -- www.tudorgirba.com www.feenk.com "Beauty is where we see it."

5 7