- Pharo-dev - lists.pharo.org

Re: [Pharo-dev] Use #shouldnt:raise: only with specific errors, or evaluate the given expression directly

by Philippe Marschall

On 24.10.13 15:24, Igor Stasenko wrote: > > > > On 24 October 2013 08:17, Philippe Marschall > <philippe.marschall(a)netcetera.ch > <mailto:philippe.marschall@netcetera.ch>> wrote: > > On 22.10.13 00:08, Camillo Bruni wrote: > > see my long explanation here > https://pharo.fogbugz.com/__default.asp?11876#87218 > <https://pharo.fogbugz.com/default.asp?11876#87218> > it looks unsuspicous until the moment you try understand such a > failing assertion. > > > This isn't moving Pharo foward. This doesn't make the system any > more flexible, adaptable, modular or malleable. This doesn't make > the system any easier to maintain. This doesn't make the system any > smaller, faster, scalable or secure. This doesn't make the system > any easier to maintain. This doesn't make life easier for anybody > developing Pharo or using Pharo. This doesn't improve Pharo in any way. > > This only adds code who's sole purpose is to break people's existing > code. > > If you disagree with such a way of writing tests then the right > solution IMHO is to write a SLint rule. > > > Running SLint over tests? I could agree the working code should be > checked for rules.. but tests? Yes. Test should have the same quality standards as the rest of the code because you'll have to maintain them just the same. Our tests have to be portable like the rest of our code. Cheers Philippe

Oct. 24, 2013

Re: [Pharo-dev] Issue #11876 (Do not allow shouldnt:raise: Error) and Seaside dev. on Pharo 3.0

by Philippe Marschall

On 24.10.13 11:59, Norbert Hartl wrote: > > Am 24.10.2013 um 10:47 schrieb Philippe Marschall > <philippe.marschall(a)netcetera.ch > <mailto:philippe.marschall@netcetera.ch>>: > >> On 24.10.13 08:05, Marcus Denker wrote: >>> >>> On 23 Oct 2013, at 23:07, StÃ©phane Ducasse >>> <stephane.ducasse(a)inria.fr >>> <mailto:stephane.ducasse@inria.fr>> wrote: >>> >>>> >>>>> Hi, >>>>> >>>>> just as info: looks like already fixed Pharo issue #11876 makes >>>>> problem for Seaside in 3.0: >>>>> >>>>> http://lists.squeakfoundation.org/pipermail/seaside-dev/2013-October/005509… >>>>> http://lists.squeakfoundation.org/pipermail/seaside-dev/2013-October/005540… >>>> >>>> >>>> It would be a bit sad that Seaside does not work in Pharo3.0 >>>> I imagine that Philippe is tired. >>>> >>>> I would revert the change and add a rule that check that the >>>> argument of shouldnt:raise: should not be an Error >>>> >>> >>> yes, people are not willing to change code all the timeâ¦ >> >> We are willing to change the code where there is a clear benefit. For >> example it's hard to argue that the new filesystem is an improvement >> over FileDirectory. We have a partically working Spec based UI for the >> servers. We would happily change the walkback to a new context API >> should it fix the bugs in the existing one. However this isn't the >> case with this change. To sole and only purpose of this change is to >> break code. This is what annoys me. >> > Personally I think the new filesystem is way better than the old one. Yeah, I meant it's better. I guess I should proof read my mails more. Cheers Philippe

Oct. 24, 2013

Re: [Pharo-dev] Use #shouldnt:raise: only with specific errors, or evaluate the given expression directly

by Igor Stasenko

On 24 October 2013 08:17, Philippe Marschall < philippe.marschall(a)netcetera.ch> wrote: > On 22.10.13 00:08, Camillo Bruni wrote: > >> see my long explanation here https://pharo.fogbugz.com/** >> default.asp?11876#87218<https://pharo.fogbugz.com/default.asp?11876#87218> >> it looks unsuspicous until the moment you try understand such a failing >> assertion. >> > > This isn't moving Pharo foward. This doesn't make the system any more > flexible, adaptable, modular or malleable. This doesn't make the system any > easier to maintain. This doesn't make the system any smaller, faster, > scalable or secure. This doesn't make the system any easier to maintain. > This doesn't make life easier for anybody developing Pharo or using Pharo. > This doesn't improve Pharo in any way. > > This only adds code who's sole purpose is to break people's existing code. > > If you disagree with such a way of writing tests then the right solution > IMHO is to write a SLint rule. > > Running SLint over tests? I could agree the working code should be checked for rules.. but tests? Useful, to some pedantic souls i guess , but not me :) But anyways, i agree with you, we should not dictate what is 'best' rules and 'how you should do the coding', we shall let people discover own and let them shoot into their own foot, if they insisting to. > Cheers > Philippe > > > -- Best regards, Igor Stasenko.

Oct. 24, 2013

[pharo-project/pharo-core]

by GitHub

Branch: refs/tags/30520 Home: https://github.com/pharo-project/pharo-core

Oct. 24, 2013

[pharo-project/pharo-core] 9e1723: 30520

by GitHub

Branch: refs/heads/3.0 Home: https://github.com/pharo-project/pharo-core Commit: 9e1723ff19630cd7c4ffda1634be0f3200e2ba64 https://github.com/pharo-project/pharo-core/commit/9e1723ff19630cd7c4ffda16… Author: Jenkins Build Server <board(a)pharo-project.org> Date: 2013-10-24 (Thu, 24 Oct 2013) Changed paths: R Compiler.package/Compiler.class/class/utilities/recompileAllFrom_.st A Kernel.package/Behavior.class/instance/accessing/localSelectors.st M KernelTests.package/BehaviorTest.class/instance/tests/testBehaviorRespectsPolymorphismWithTraitBehavior.st M Morphic-Base.package/MorphTreeMorph.class/instance/column handling/headerBounds.st M Morphic-Base.package/MorphTreeMorph.class/instance/submorphs-add%2Fremove/addMorphsAfter_fromCollection_.st M Morphic-Base.package/MorphTreeNodeMorph.class/instance/accessing/spacerWidth.st M Morphic-Base.package/MorphTreeNodeMorph.class/instance/layout/layoutBounds.st R OpalCompiler-Core.package/ManifestOpalCompilerCore.class/class/meta data/ruleCodeCruftLeftInMethodsRuleV1FalsePositive.st A ScriptLoader30.package/ScriptLoader.class/instance/pharo - scripts/script175.st A ScriptLoader30.package/ScriptLoader.class/instance/pharo - updates/update30520.st M ScriptLoader30.package/ScriptLoader.class/instance/public/commentForCurrentUpdate.st Log Message: ----------- 30520 11982 testExplicitRequirementMethodsShouldBeImp... https://pharo.fogbugz.com/f/cases/11982 11991 remove #recompileAllFrom: https://pharo.fogbugz.com/f/cases/11991 11988 MorphTreeMorph incorrect bounds https://pharo.fogbugz.com/f/cases/11988 http://files.pharo.org/image/30/30520.zip

Oct. 24, 2013

Re: [Pharo-dev] Startup Preferences: An Alternative

by Igor Stasenko

On 24 October 2013 14:42, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > And to save on parenthesis, you could also write > > Compiler evaluate: 'http://my.url.with.script.st' asZnUrl > retrieveContents. > > CodeImporter evaluateString: 'http://my.url.with.script.st' asUrl > retrieveContents. > > this is why i love smalltalk: that you can express highly complex stuff (there's a lot of stuff happens under the hood) by single line expression. Comparing to most examples of C frameworks, where you had to manually allocate buffers, initialize some data structures, and pass several options and flags here and there, which at the end does what you need, but floods your mind with hundreds of tiny irrelevant details.. > With #asUrl and #asZnUrl soon to be 100% the same thing. > > On 24 Oct 2013, at 14:32, Esteban Lorenzano <estebanlm(a)gmail.com> wrote: > > > you mean: > > > > CodeImporter evaluateString: (ZnEasy get: 'my-url') contents. > > > > :) > > > > On Oct 23, 2013, at 11:17 PM, Igor Stasenko <siguctua(a)gmail.com> wrote: > > > >> > >> On 23 October 2013 20:21, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > >> This is really cool, is that in the image already ? > >> What about plain HTTP repos ? > >> > >> > >> indeed, what could be simpler than: > >> > >> Compiler evaluate: (ZnEasy get: 'http://my.url.with.script.st') > >> > >> and you can, of course place it on github, if you need version control, > like: > >> > >> > https://raw.github.com/pharo-project/pharo-vm/master/codegen-scripts/LoadVM… > >> > >> > >> On 23 Oct 2013, at 20:04, Sean P. DeNigris <sean(a)clipperadams.com> > wrote: > >> > >> > philippeback wrote > >> >> Would be great to have a logout, in order to give images to other > people > >> >> without leaving credentials around. > >> > > >> > Done, added to world menu under the individual repo login options > >> > > >> > The way the logins work is that a file named .mcconfig is expected in > the > >> > general preferences folder with the format: > >> > > >> > [squeaksource] > >> > user := 'me'. > >> > password := 'pwd'. > >> > [squeaksource3] > >> > user := 'me'. > >> > password = 'pwd'. > >> > [smalltalkhub] > >> > user := 'me'. > >> > password := 'pwd'. > >> > > >> > > >> > > >> > ----- > >> > Cheers, > >> > Sean > >> > -- > >> > View this message in context: > http://forum.world.st/Startup-Preferences-An-Alternative-tp4713968p4716686.… > >> > Sent from the Pharo Smalltalk Developers mailing list archive at > Nabble.com. > >> > > >> > >> > >> > >> > >> > >> -- > >> Best regards, > >> Igor Stasenko. > > > > > -- Best regards, Igor Stasenko.

Oct. 24, 2013

Re: [Pharo-dev] Some very rough benchmarks on text rendering

by Nicolas Cellier

Note that Squeak is running the display twice with StrikeFont when subPixelRenderColorFonts is enabled, which is far from optimal. See BitBlt>>displayString:from:to:at:strikeFont:kern: and BitBlt>>installStrikeFont:foregroundColor:backgroundColor: 2013/10/24 phil(a)highoctane.be <phil(a)highoctane.be> > Very cool indeed. > > The days of fearing to click on large methods due to looong rendering > times are numbered (e.g. Seaside FileLibrary uploaded files... like jQuery > or images). > > PluggableTextMorph does a hell of a lot of twisted things, I guess we'll > get much improved understandability as well with the Tx things. > > Kepp up the good work, it is much appreciated! > > Phil > > > On Thu, Oct 24, 2013 at 8:41 AM, ClÃ©ment Bera <bera.clement(a)gmail.com>wrote: > >> Awesome ! >> >> I'd love to have that in the image and use it for my everyday dev. >> Great work Igor please finish it so we can use it :) >> >> >> 2013/10/24 Igor Stasenko <siguctua(a)gmail.com> >> >>> As a followup on discussion about improving/benching text stuff. >>> Just out of curiosity, wanted to compare how sloooow athens with cairo >>> comparing to good old bitblt in terms of displaying text. >>> >>> See for yourselves. >>> Sure comparison is very rough (text are different >>> - in athens i using freetype font, in workspace is raster font) >>> but at least both windows displaying same amount of text. >>> (that i think what is relevant to us, humans ;) ) >>> And, of course, to make things more precise need to write more code to >>> make sure >>> it renders exact contents of the text morph fully without any clipping, >>> in both cases). >>> And frankly, i am not in a mood of making/coding precise benchmark now. >>> :) >>> >>> Why i posting this, is mainly because i am quite surprised by numbers. >>> i expected opposite proportion.. >>> >>> And i still almost didn't optimized things. >>> >>> [image: Inline images 2][image: Inline images 1] >>> >>> -- >>> Best regards, >>> Igor Stasenko. >>> >> >> >

Oct. 24, 2013

Re: [Pharo-dev] Startup Preferences: An Alternative

by Sven Van Caekenberghe

And to save on parenthesis, you could also write Compiler evaluate: 'http://my.url.with.script.st' asZnUrl retrieveContents. CodeImporter evaluateString: 'http://my.url.with.script.st' asUrl retrieveContents. With #asUrl and #asZnUrl soon to be 100% the same thing. On 24 Oct 2013, at 14:32, Esteban Lorenzano <estebanlm(a)gmail.com> wrote: > you mean: > > CodeImporter evaluateString: (ZnEasy get: 'my-url') contents. > > :) > > On Oct 23, 2013, at 11:17 PM, Igor Stasenko <siguctua(a)gmail.com> wrote: > >> >> On 23 October 2013 20:21, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: >> This is really cool, is that in the image already ? >> What about plain HTTP repos ? >> >> >> indeed, what could be simpler than: >> >> Compiler evaluate: (ZnEasy get: 'http://my.url.with.script.st') >> >> and you can, of course place it on github, if you need version control, like: >> >> https://raw.github.com/pharo-project/pharo-vm/master/codegen-scripts/LoadVM… >> >> >> On 23 Oct 2013, at 20:04, Sean P. DeNigris <sean(a)clipperadams.com> wrote: >> >> > philippeback wrote >> >> Would be great to have a logout, in order to give images to other people >> >> without leaving credentials around. >> > >> > Done, added to world menu under the individual repo login options >> > >> > The way the logins work is that a file named .mcconfig is expected in the >> > general preferences folder with the format: >> > >> > [squeaksource] >> > user := 'me'. >> > password := 'pwd'. >> > [squeaksource3] >> > user := 'me'. >> > password = 'pwd'. >> > [smalltalkhub] >> > user := 'me'. >> > password := 'pwd'. >> > >> > >> > >> > ----- >> > Cheers, >> > Sean >> > -- >> > View this message in context: http://forum.world.st/Startup-Preferences-An-Alternative-tp4713968p4716686.… >> > Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com. >> > >> >> >> >> >> >> -- >> Best regards, >> Igor Stasenko. >

Oct. 24, 2013

Re: [Pharo-dev] Startup Preferences: An Alternative

by Esteban Lorenzano

you mean: CodeImporter evaluateString: (ZnEasy get: 'my-url') contents. :) On Oct 23, 2013, at 11:17 PM, Igor Stasenko <siguctua(a)gmail.com> wrote: > > On 23 October 2013 20:21, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > This is really cool, is that in the image already ? > What about plain HTTP repos ? > > > indeed, what could be simpler than: > > Compiler evaluate: (ZnEasy get: 'http://my.url.with.script.st') > > and you can, of course place it on github, if you need version control, like: > > https://raw.github.com/pharo-project/pharo-vm/master/codegen-scripts/LoadVM… > > > On 23 Oct 2013, at 20:04, Sean P. DeNigris <sean(a)clipperadams.com> wrote: > > > philippeback wrote > >> Would be great to have a logout, in order to give images to other people > >> without leaving credentials around. > > > > Done, added to world menu under the individual repo login options > > > > The way the logins work is that a file named .mcconfig is expected in the > > general preferences folder with the format: > > > > [squeaksource] > > user := 'me'. > > password := 'pwd'. > > [squeaksource3] > > user := 'me'. > > password = 'pwd'. > > [smalltalkhub] > > user := 'me'. > > password := 'pwd'. > > > > > > > > ----- > > Cheers, > > Sean > > -- > > View this message in context: http://forum.world.st/Startup-Preferences-An-Alternative-tp4713968p4716686.… > > Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com. > > > > > > > > -- > Best regards, > Igor Stasenko.

Oct. 24, 2013

Re: [Pharo-dev] Issue #11876 (Do not allow shouldnt:raise: Error) and Seaside dev. on Pharo 3.0

by Sven Van Caekenberghe

On 24 Oct 2013, at 13:55, Torsten Bergmann <astares(a)gmx.de> wrote: > From the comments I think we can agree that it is a better to have a lint rule > instead of breaking things the way it was introduced by the original issue #11876. > > I therefore opened an issue > https://pharo.fogbugz.com/f/cases/11989/shouldnt-raise-Error-from-issue-118… > and uploaded a slice > > SLICE-Issue-11989-shouldntraise-Error-from-issue-11876-is-too-offensive-TorstenBergmann.1 > > to Pharo30Inbox on STHub (http://smalltalkhub.com/mc/Pharo/Pharo30Inbox/main) that > removed #validateShouldntException: > > I used a fresh 3.0 image (Pharo3.0 Latest update: #30519) > > Please test so it can be integrated and we can continue to focus on more interesting stuff. With respect to Seaside (and its cross-platform-ness), this is probably a good decision, for now. But it is fundamentally unfair to call this an arbitrary unimportant change meant to annoy people. On the contrary ! I even think that there are too many methods in TAssertable and the #shouldnt: checks are very confusing and are candidates for removal, IMHO. Too much unnecessary code is not good in the long run. Have you ever written a #shouldnt: test ? I have never done so ! I took a Seaside image and looked at the senders of #shouldnt: and friends. There are that many from Seaside and I havenât seen any good example that could not be written otherwise. My 2c. Sven > Thanks > Torsten > > >

Oct. 24, 2013