- Pharo-dev - lists.pharo.org

Integer>>#putOn:
by Esteban Lorenzano Oct. 12, 2015

Oct. 12, 2015

Hi, I just spotted this: Integer>>#putOn: aStream (aStream isBinary ifTrue: [ self asByteArray ] ifFalse: [ self asString]) putOn: aStream It does not look so bad at the begining, butâ¦ if I do: ByteArray streamContents: [ :s | 42 putOn: s ] I got an infinite loop. I know, it does not look as a very typical case, butâ¦ if Iâm writing to a Socket and I want to use mySocket << 255 âIACâ, thatâs exactly what will happen :( If I redefine it as: Integer>>#putOn: aStream aStream isBinary ifFalse: [ self asString putOn: aStream. ^ self ]. self asByteArray do: [ :each | aStream nextPut: each ] then is okâ¦ Would that be a good change? Esteban

2 2

seg fault on Ubuntu when launching
by Robert Withers Oct. 12, 2015

Oct. 12, 2015

Hello, I am new to Pharo and just getting into it. I have Ubuntu and used PPA to install Pharo. The launcher is failing with a segfault crash.dmp when trying to launch. It is the UnixOSProcessAccessor>>forkAndExec... that is segaulting. I looked in /usr/lib/pharo-vm and do not see a UnixOSProcess plugin. Is it internal? I am able to launch an image from outside the launcher. -- thanks, Robert

2 4

FilePlugin primitiveFail on Win7 slave
by Max Leske Oct. 12, 2015

Oct. 12, 2015

Hey guys, is there a known problem with the FilePlugin on the Win7 slave? Whenever I try to delete a file I get a primitiveFailed. As you can see in the trace below, Iâm actually even checking for FileNotFound, which means that the file exists but the VM canât perform the operation for some reason. Cheers, Max FilePluginPrims(Object)>>primitiveFailed: FilePluginPrims(Object)>>primitiveFailed FilePluginPrims>>deleteDirectory: WindowsStore(DiskStore)>>delete: FileSystem>>delete: FileReference>>delete DeleteVisitor>>visitReference: DeleteVisitor(FileSystemVisitor)>>visitDirectory: PostorderGuide>>visit: Message>>sendTo: [ self pop sendTo: self ] in PostorderGuide>>show: PostorderGuide(FileSystemGuide)>>whileNotDoneDo: PostorderGuide>>show: PostorderGuide class(FileSystemGuide class)>>show:to:selecting: DeleteVisitor>>visit: DeleteVisitor class>>delete: FileReference(AbstractFileReference)>>deleteAll [ ref deleteAll ] in [ :ref | ref exists ifFalse: [ ^ self ] ifTrue: [ [ ref deleteAll ] on: FileDoesNotExist do: [ ] "ignore. Seems to be a glitch in the File plugin" ] ] in LGitTestRepository class>>tearDown BlockClosure>>on:do: [ :ref | ref exists ifFalse: [ ^ self ] ifTrue: [ [ ref deleteAll ] on: FileDoesNotExist do: [ ] "ignore. Seems to be a glitch in the File plugin" ] ] in LGitTestRepository class>>tearDown FileReference(Object)>>in: LGitTestRepository class>>tearDown LGitBlobTest(LGitTestCase)>>tearDown

2 2

[pharo-project/pharo-core]
by GitHub Oct. 12, 2015

Oct. 12, 2015

Branch: refs/tags/50381 Home: https://github.com/pharo-project/pharo-core

1 0

[pharo-project/pharo-core] b3031d: 50381
by GitHub Oct. 12, 2015

Oct. 12, 2015

Branch: refs/heads/5.0 Home: https://github.com/pharo-project/pharo-core Commit: b3031d243c04ad2bd5c673f9805499c4852e4661 https://github.com/pharo-project/pharo-core/commit/b3031d243c04ad2bd5c673f9… Author: Jenkins Build Server <board(a)pharo-project.org> Date: 2015-10-12 (Mon, 12 Oct 2015) Changed paths: R Collections-Abstract.package/Collection.class/instance/accessing/atRandom.st R Collections-Abstract.package/Collection.class/instance/accessing/atRandom_.st R Collections-Abstract.package/SequenceableCollection.class/instance/accessing/atRandom_.st R Collections-Tests.package/SetTest.class/instance/some tests/testAtRandom.st R Collections-Unordered.package/Matrix.class/instance/accessing/atRandom.st R Collections-Unordered.package/Matrix.class/instance/accessing/atRandom_.st R Collections-Unordered.package/Set.class/instance/accessing/atRandom_.st R Kernel-Tests.package/RandomTest.class/README.md R Kernel-Tests.package/RandomTest.class/definition.st R Kernel-Tests.package/RandomTest.class/instance/running/setUp.st R Kernel-Tests.package/RandomTest.class/instance/tests/testNext.st M Kernel.package/Context.class/class/instance creation/basicNew_.st M Kernel.package/Context.class/instance/controlling/shortDebugStack.st M Kernel.package/Context.class/instance/printing/debugStackOn_.st M Kernel.package/Context.class/instance/printing/debugStack_on_.st M Kernel.package/Context.class/instance/printing/printOn_.st M Kernel.package/Exception.class/instance/handling/freeze.st R Kernel.package/Integer.class/instance/truncation and round off/atRandom.st R Kernel.package/Integer.class/instance/truncation and round off/atRandom_.st R Kernel.package/Random.class/README.md R Kernel.package/Random.class/class/instance creation/seed_.st R Kernel.package/Random.class/class/testing/bucketTest_.st R Kernel.package/Random.class/class/testing/theItsCompletelyBrokenTest.st R Kernel.package/Random.class/definition.st R Kernel.package/Random.class/instance/accessing/next.st R Kernel.package/Random.class/instance/accessing/nextInt_.st R Kernel.package/Random.class/instance/accessing/next_.st R Kernel.package/Random.class/instance/accessing/next_into_.st R Kernel.package/Random.class/instance/initialization/initialize.st R Kernel.package/Random.class/instance/initialization/seed_.st R Kernel.package/Random.class/instance/private/nextValue.st R Kernel.package/Random.class/instance/private/seed.st A Random.package/Random.class/README.md A Random.package/Random.class/class/instance creation/seed_.st A Random.package/Random.class/class/testing/bucketTest_.st A Random.package/Random.class/definition.st A Random.package/Random.class/instance/accessing/next.st A Random.package/Random.class/instance/accessing/nextInt_.st A Random.package/Random.class/instance/accessing/next_.st A Random.package/Random.class/instance/accessing/next_into_.st A Random.package/Random.class/instance/initialization/initialize.st A Random.package/Random.class/instance/initialization/seed_.st A Random.package/Random.class/instance/private/privateNextSeed.st A Random.package/Random.class/instance/private/privateNextValue.st A Random.package/Random.class/instance/private/seed.st A Random.package/RandomTest.class/README.md A Random.package/RandomTest.class/definition.st A Random.package/RandomTest.class/instance/running/setUp.st A Random.package/RandomTest.class/instance/tests/testIfCompletelyBroken.st A Random.package/RandomTest.class/instance/tests/testNext.st A Random.package/SharedRandom.class/README.md A Random.package/SharedRandom.class/class/accessing/globalGenerator.st A Random.package/SharedRandom.class/class/initialization/initialize.st A Random.package/SharedRandom.class/definition.st A Random.package/SharedRandom.class/instance/accessing/next.st A Random.package/SharedRandom.class/instance/accessing/nextInt_.st A Random.package/SharedRandom.class/instance/accessing/next_into_.st A Random.package/SharedRandom.class/instance/initialization/initialize.st A Random.package/extension/Collection/instance/atRandom.st A Random.package/extension/Collection/instance/atRandom_.st A Random.package/extension/Integer/instance/atRandom.st A Random.package/extension/Integer/instance/atRandom_.st A Random.package/extension/Matrix/instance/atRandom.st A Random.package/extension/Matrix/instance/atRandom_.st A Random.package/extension/SequenceableCollection/instance/atRandom_.st A Random.package/extension/Set/instance/atRandom_.st A Random.package/extension/SetTest/instance/testAtRandom.st R ScriptLoader50.package/ScriptLoader.class/instance/pharo - scripts/script50380.st A ScriptLoader50.package/ScriptLoader.class/instance/pharo - scripts/script50381.st R ScriptLoader50.package/ScriptLoader.class/instance/pharo - updates/update50380.st A ScriptLoader50.package/ScriptLoader.class/instance/pharo - updates/update50381.st M ScriptLoader50.package/ScriptLoader.class/instance/public/commentForCurrentUpdate.st Log Message: ----------- 50381 16761 Method source should not contain linefeeds in Kernel https://pharo.fogbugz.com/f/cases/16761 16577 Random should be in its own package https://pharo.fogbugz.com/f/cases/16577 http://files.pharo.org/image/50/50381.zip

1 0

MultistateLabelMorph still required
by Torsten Bergmann Oct. 12, 2015

Oct. 12, 2015

Hi, from Web history I see that it was a Polymorph extension from Gary Chambers: "Added MultistateLabelMorph that can adapt to interaction sates from buttons." http://forum.world.st/update-1-1-11293-Polymorph-Enhancements-td1680386.html In Pharo5.0 Latest update: #50380 one can see that it is not referenced. Do we still need it or can we move it to deprecated? Thanks T.

2 1

Re: [Pharo-dev] Yet another Athens font thread
by Aliaksei Syrel Oct. 11, 2015

Oct. 11, 2015

No, they are needed. Because glyphs must be calculated (getGlyphs:of:) after(!) setting font size. Otherwise there are huge gaps between characters. When font problem appears, some characters are of size 14 even if font size 10. So glyphs of big characters have larger extent as it should be. In your example you set font size and fix font, so all characters are rendered correctly, but at wrong positions. That is a source of unnecessary gaps. Thete is a way to break font: 1) Get new pharo image 2) Get Bloc 3) Switch to bloc 4) remove all morphs 5) save and quite 6) open image 7) open nautilus Do something to fix font. Repeat 4-7. If fix is "fake" fonts will be broken.. On Oct 9, 2015 10:31 PM, "Nicolai Hess" <nicolaihess(a)web.de> wrote: > > > 2015-10-08 18:05 GMT+02:00 Aliaksei Syrel <alex.syrel(a)gmail.com>: > >> Do you remember a hack with scaling? (scaleBy: 1.0001(a)1.0001) It makes >> letters to have the same height, so no more huge ones with small fonts. But >> it doesn't fix a bug when letters have wrong width so there is a space >> between characters. The same effect can be achieved if we set font size >> manually after setting scaled font before drawing glyphs with cairo. It >> means that at some point we get wrong calculated glyphs. So if we "fix" >> scaled font before glyph calculation - the problem would be gone. In the >> patch I set scaled font, manually apply font size, retrieve "fixed" scaled >> font and calculate glyphs based on it. >> >> Also after investigation I found something interesting. If we clean real >> font of a logical font the problem gets fixed for some period of time. I >> assume that FreeTypeFont gets broken after a while and logical font always >> returns broken real font. Setting realFont to nil forces logical font to >> create a new working free type font. As soon as new free type font is >> created cairo produces scaled font and caches it. So actual problem lies >> deep in FreeTypeFont... >> >> I even opened two images: one with broken fonts and one with working. >> Then I compared in inspector two free type fonts - they are absolutely the >> same and have identical instance variable in a whole object tree, except of >> handler pointers to C structure which I can't inspect from pharo... >> >> Saving image fixes fonts because of >> >>> LogicalFont>>shutDown: quitting >>> self allSubInstances do: [:i | i clearRealFont]. >> >> >> Cheers, >> Alex >> > > This is great. > > I looked at your change. And noticed the call to > setFontSize > this was never called until now. Maybe this was the problem. > > Now, I just changed > > renderGlyphs: cairoGlyphs > > font lock. > canvas > setPathMatrix; > setScaledFont: font; > setFontSize: originalFont pixelSize. "<< new " > canvas paint loadOnCairoCanvas: canvas. > > canvas showGlyphs: cairoGlyphs address size: cairoGlyphs size. > font unlock. > > And it seems the font issue is gone. Maybe the other changes (set and get > scaled font in #glyphsOf:from:to: ) > isn't needed? (needs more tests) > > > >> On Thu, Oct 8, 2015 at 5:43 PM, Nicolai Hess <nicolaihess(a)web.de> wrote: >> >>> How did you fix it? >>> Am 08.10.2015 17:06 schrieb "Aliaksei Syrel" <alex.syrel(a)gmail.com>: >>> >>>> Hi >>>> >>>> It looks like I fixed that weird text rendering problem... Just get the >>>> latest Athens-Cairo. >>>> >>>> Cheers, >>>> Alex >>>> >>>> On Mon, Aug 3, 2015 at 9:32 PM, Nicolai Hess <nicolaihess(a)web.de> >>>> wrote: >>>> >>>>> >>>>> >>>>> 2015-06-10 12:01 GMT+02:00 Aliaksei Syrel <alex.syrel(a)gmail.com>: >>>>> >>>>>> Hi >>>>>> >>>>>> I noticed that if I force use "Source Code Pro" as standard font for >>>>>> all elements in the image I never have problems with font rendering. In >>>>>> case of "Source Sans Pro" font is just not usable becasue of rendering bugs >>>>>> on mac, windows and linux(?). If "Lucida Sans Pro" is used - no bugs too. >>>>>> >>>>>> Because bug is reproducable on all platform - it's platform >>>>>> independent bug. >>>>>> Because it's reproducable only with one specific font - maybe it's >>>>>> not an athens issue but bug in font itself and if someone could >>>>>> change/update Source Sans Pro to a newly version or to a similar font the >>>>>> problem will be solved? >>>>>> >>>>> >>>>> Did you try to use a new font version ? >>>>> I don't know exactly how embedded fonts are included. I tried to >>>>> create the font with a newer version but it did not work ( or at least I >>>>> could not see any difference). >>>>> >>>>> >>>>> >>>>>> >>>>>> Cheers, >>>>>> Alex >>>>>> >>>>> >>>>> >>>> >> >

2 1

Re: [Pharo-dev] hash and Collections. A solution for some.
by Ralph Boland Oct. 11, 2015

Oct. 11, 2015

I haven't followed this thread closely so I apologize in advance if I am somewhat off topic but I think what I have to say may be of interest to some of you. I think the current method of determining a hash value for a collection is fine in the general case and if it is inadequate for your particular needs then implement your own collection class that does what you want. This was the case for one of my applications so I created classes HashArray and HashSet. Both of these classes used as their hash value the XOR of the the hash values of the collection elements (which was of course stored in an instance variable). Each time an element was added, removed, or replaced in the collection the hash value of the collection had to be recomputed (required O(1) time). Of course this approach breaks down if the collection elements can change their hash values. My solution is not ideal for the hashArray case because two hashArrays containing the same elements but in a different order would have the same hash value. But this didn't matter for my application because I used hashArrays as a set converted into an array (with values sorted) for time/space efficiency reasons. This worked really well for me, in part because the elements stored in the hashArrays and hashSets also had really good hash values. Also note that, in my application, two hashArrays or hashSets containing the same elements were considered to be equal. If anyone makes use of this idea I wouldn't mind a quick email on your experience. Ralph Boland

1 0

A really cool thread on fog bug
by stepharo Oct. 11, 2015

Oct. 11, 2015

https://pharo.fogbugz.com/f/cases/16577/Random-should-be-in-its-own-package I love the spirit of this thread!!! Thank you for this example of collaboration. I wrote : I want to **really** thank you. This thread is great and show the power of collaboration. We all stronger and better. Thanks again.

2 1

hash and Collections
by Esteban Lorenzano Oct. 11, 2015

Oct. 11, 2015

Hi, Can anyone explain me how this is correct? Collection>>#hash "Answer an integer hash value for the receiver such that, -- the hash value of an unchanged object is constant over time, and -- two equal objects have equal hash values" | hash | hash := self species hash. self size <= 10 ifTrue: [self do: [:elem | hash := hash bitXor: elem hash]]. ^hash bitXor: self size hash I meanâ¦ a hash that changes when I add elements to the collectionâ¦ how can this work? Esteban

8 11