January 2013 - Pharo-dev - lists.pharo.org

Re: [Pharo-project] DocTest

by Tudor Girba

FYI, we are already using Phexample in Moose since several years, and it just works out of the box next to Sunit. Doru On Wednesday, January 23, 2013, Sven Van Caekenberghe wrote: > Yes we #should ;-) > > It is beautiful. > > On 23 Jan 2013, at 11:04, Camillo Bruni <camillobruni(a)gmail.com<javascript:;>> > wrote: > > > check the project page for all examples and explanations: > > > > > http://smalltalkthoughts.blogspot.ch/2009/11/phexample-because-examples-exp… > > http://www.iam.unibe.ch/~akuhn/blog/2009/shoulda-use-this-in-pharo > > > > to me this feels more like Smalltalk and less like Java ;) > > > > On 2013-01-23, at 08:47, Nicolas Cellier < > nicolas.cellier.aka.nice(a)gmail.com <javascript:;>> wrote: > >> Who should assert then? > >> From a certain POV, it does not seem that awfully unlogical to > >> delegate this responsability to a TestCase. > >> A TestCase is here to assert some rule. > >> > >> Nicolas > >> > >> 2013/1/23 Camillo Bruni <camillobruni(a)gmail.com <javascript:;>>: > >>> for that we already have a fully working Phexample: > >>> > >>> http://www.squeaksource.com/phexample.html > >>> > >>> which I will push in 3.0, cause I never really liked #assert: and > friends being > >>> sent to self. Furthermore we could finally build up decent messages! > >>> > >>> On 2013-01-22, at 18:24, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr<javascript:;>> > wrote: > >>> > >>>> it would be nice to introduce the same syntax than Sunit > >>>> > >>>> exp assert: exp2 > >>>> > >>>> [ exp ] shouldraise: > >>>> > >>>> Stef > >>>> > >>>> > >>>> On Jan 21, 2013, at 4:41 AM, Benjamin wrote: > >>>> > >>>>> Hi guys, > >>>>> > >>>>> I am attending a python lecture and discovered DocTest[1], and I was > wondering if such a tool exists for Pharo ? :) > >>>>> > >>>>> > >>>>> Ben > >>>>> > >>>>> [1]http://docs.python.org/2/library/doctest.html > >>>>> > >>>>> > >>>> > >>>> > >>> > >>> > >> > > > > > > > -- www.tudorgirba.com "Every thing has its own flow"

Jan. 23, 2013

Re: [Pharo-project] [Pharo-users] Pharo 2.0 - System Error Handling Failed

by Marcus Denker

On Jan 23, 2013, at 10:52 AM, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > I really think that for the process of polishing 2.0 from beta to release we need a new label for issues. > > Bugs like the one below are really important, they might be silly or easy to fix, but they leave a bad impression. We want as many people as possible to help, but UX failures like this will scare away many users, most won't be as helpful as Andy to even report the issue. We as developers know that 2.0 is OK because we use it every day, but occasional users who just try 2.0 for the first time can't know that. > > So not HorribleErrorBlockingEverything but maybe HorribleUXQualityBugScaringAwayNormalPeople as label ? > :-) yes. Marcus

Jan. 23, 2013

Re: [Pharo-project] DocTest

by Sven Van Caekenberghe

Yes we #should ;-) It is beautiful. On 23 Jan 2013, at 11:04, Camillo Bruni <camillobruni(a)gmail.com> wrote: > check the project page for all examples and explanations: > > http://smalltalkthoughts.blogspot.ch/2009/11/phexample-because-examples-exp… > http://www.iam.unibe.ch/~akuhn/blog/2009/shoulda-use-this-in-pharo > > to me this feels more like Smalltalk and less like Java ;) > > On 2013-01-23, at 08:47, Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com> wrote: >> Who should assert then? >> From a certain POV, it does not seem that awfully unlogical to >> delegate this responsability to a TestCase. >> A TestCase is here to assert some rule. >> >> Nicolas >> >> 2013/1/23 Camillo Bruni <camillobruni(a)gmail.com>: >>> for that we already have a fully working Phexample: >>> >>> http://www.squeaksource.com/phexample.html >>> >>> which I will push in 3.0, cause I never really liked #assert: and friends being >>> sent to self. Furthermore we could finally build up decent messages! >>> >>> On 2013-01-22, at 18:24, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: >>> >>>> it would be nice to introduce the same syntax than Sunit >>>> >>>> exp assert: exp2 >>>> >>>> [ exp ] shouldraise: >>>> >>>> Stef >>>> >>>> >>>> On Jan 21, 2013, at 4:41 AM, Benjamin wrote: >>>> >>>>> Hi guys, >>>>> >>>>> I am attending a python lecture and discovered DocTest[1], and I was wondering if such a tool exists for Pharo ? :) >>>>> >>>>> >>>>> Ben >>>>> >>>>> [1]http://docs.python.org/2/library/doctest.html >>>>> >>>>> >>>> >>>> >>> >>> >> > >

Jan. 23, 2013

Re: [Pharo-project] DocTest

by Frank Shearar

Well, of course, JUnit feels that way because of SUnit, but sure. RSpec-like testing languages read much more easily than #assert:, in my not so humble opinion. frank On 23 January 2013 10:04, Camillo Bruni <camillobruni(a)gmail.com> wrote: > check the project page for all examples and explanations: > > http://smalltalkthoughts.blogspot.ch/2009/11/phexample-because-examples-exp… > http://www.iam.unibe.ch/~akuhn/blog/2009/shoulda-use-this-in-pharo > > to me this feels more like Smalltalk and less like Java ;) > > On 2013-01-23, at 08:47, Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com> wrote: >> Who should assert then? >> From a certain POV, it does not seem that awfully unlogical to >> delegate this responsability to a TestCase. >> A TestCase is here to assert some rule. >> >> Nicolas >> >> 2013/1/23 Camillo Bruni <camillobruni(a)gmail.com>: >>> for that we already have a fully working Phexample: >>> >>> http://www.squeaksource.com/phexample.html >>> >>> which I will push in 3.0, cause I never really liked #assert: and friends being >>> sent to self. Furthermore we could finally build up decent messages! >>> >>> On 2013-01-22, at 18:24, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: >>> >>>> it would be nice to introduce the same syntax than Sunit >>>> >>>> exp assert: exp2 >>>> >>>> [ exp ] shouldraise: >>>> >>>> Stef >>>> >>>> >>>> On Jan 21, 2013, at 4:41 AM, Benjamin wrote: >>>> >>>>> Hi guys, >>>>> >>>>> I am attending a python lecture and discovered DocTest[1], and I was wondering if such a tool exists for Pharo ? :) >>>>> >>>>> >>>>> Ben >>>>> >>>>> [1]http://docs.python.org/2/library/doctest.html >>>>> >>>>> >>>> >>>> >>> >>> >> > >

Jan. 23, 2013

Re: [Pharo-project] zero conf build

by Camillo Bruni

so yeah, looks like the linux VMs won't run anymore under ubuntu 10 you will can find out the last version of GLIBC on your distro by running strings /lib/libc.so.6 | grep GLIBC On 2013-01-22, at 17:41, Tudor Girba <tudor(a)tudorgirba.com> wrote: > For example: > > curl http://pharo.gforge.inria.fr/ci/script/ciPharo20PharoVM.sh | sh > ./vm.sh > > or: > > curl http://pharo.gforge.inria.fr/ci/script/ciPharo20CogVM.sh | bash. > ./vm.sh > > Cheers, > Doru

Jan. 23, 2013

Re: [Pharo-project] DocTest

by Camillo Bruni

check the project page for all examples and explanations: http://smalltalkthoughts.blogspot.ch/2009/11/phexample-because-examples-exp… http://www.iam.unibe.ch/~akuhn/blog/2009/shoulda-use-this-in-pharo to me this feels more like Smalltalk and less like Java ;) On 2013-01-23, at 08:47, Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com> wrote: > Who should assert then? > From a certain POV, it does not seem that awfully unlogical to > delegate this responsability to a TestCase. > A TestCase is here to assert some rule. > > Nicolas > > 2013/1/23 Camillo Bruni <camillobruni(a)gmail.com>: >> for that we already have a fully working Phexample: >> >> http://www.squeaksource.com/phexample.html >> >> which I will push in 3.0, cause I never really liked #assert: and friends being >> sent to self. Furthermore we could finally build up decent messages! >> >> On 2013-01-22, at 18:24, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: >> >>> it would be nice to introduce the same syntax than Sunit >>> >>> exp assert: exp2 >>> >>> [ exp ] shouldraise: >>> >>> Stef >>> >>> >>> On Jan 21, 2013, at 4:41 AM, Benjamin wrote: >>> >>>> Hi guys, >>>> >>>> I am attending a python lecture and discovered DocTest[1], and I was wondering if such a tool exists for Pharo ? :) >>>> >>>> >>>> Ben >>>> >>>> [1]http://docs.python.org/2/library/doctest.html >>>> >>>> >>> >>> >> >> >

Jan. 23, 2013

Re: [Pharo-project] OrderedCollections and File I/O

by Camillo Bruni

which reminds me, we should have a dumb default serializer method on Object: Object new serializeTo: 'aFile' asFileReference cause I never figure out which Fuel class to use for serialization :) On 2013-01-23, at 08:54, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > On 23 Jan 2013, at 05:24, J. P. L. MartÃn <arguser(a)gmail.com> wrote: > >> Say I have an OrderedCollection which I want to save into a file and then recover from that file when I restart the application. How should I proceed? So far I've been doing a mess, saving the OrderedCollection on a file asString, since I'm then unable to recover it asOrderedCollection. >> >> Another problem that I have is at the moment of showing items from that OrderedCollection on a MultiColumnListMorph where I get unwanted info from the printOn: >> >> a Prestamo fechaPrevistaDevolucion: 20 January 2013 fechaPrestamo: 17 January 2013 id: ''1'' idSocio: ''2'' idEjemplarl: ''1'' fechaDevolucion: ''''' >> >> Where I want: >> >> 20 January 2013 17 January 2013 1 2 1 '' >> >> Here you can find the source code: >> https://github.com/arguser/Paradigmas-Final > > Fuel is a fast binary serializer that can store almost anything. > > If the collection is holding clean domain level objects, you could also use XML, JSON or STON. > These give you human readable textual representations. > > Sven > > Search for > XML Support > NeoJSON > STON > > -- > Sven Van Caekenberghe > http://stfx.eu > Smalltalk is the Red Pill > >

Jan. 23, 2013

Re: [Pharo-project] [Pharo-users] Pharo 2.0 - System Error Handling Failed

by Sven Van Caekenberghe

I really think that for the process of polishing 2.0 from beta to release we need a new label for issues. Bugs like the one below are really important, they might be silly or easy to fix, but they leave a bad impression. We want as many people as possible to help, but UX failures like this will scare away many users, most won't be as helpful as Andy to even report the issue. We as developers know that 2.0 is OK because we use it every day, but occasional users who just try 2.0 for the first time can't know that. So not HorribleErrorBlockingEverything but maybe HorribleUXQualityBugScaringAwayNormalPeople as label ? Sven On 23 Jan 2013, at 02:43, Camillo Bruni <camillobruni(a)gmail.com> wrote: > that is/was a bug: > > https://code.google.com/p/pharo/issues/detail?id=6608 > > looks like it is still there in a newer image! > > On 2013-01-23, at 02:05, Andy Burnett <andy.burnett(a)knowinnovation.com> wrote: > >> I am running the Pharo 2.0 beta, and I am getting a lot of System Error >> Handling Failed events. Is this simply a product of the beta status, or >> have I somehow enabled some kind of debug switch? >> >> For example, in a workspace I execute a doit on: >> >> fd := FileDialogWindow new. >> fn := fd openInWorld. >> >> When I click OK in the dialog box, I get the System Error screen. On other >> occasions I get the screen, press space to continue, then I get the debug >> window, and when attempting to debug it bounces me back to the System Error >> screen. >> >> Normal behaviour for the beta, or operator error on my part? >> >> Cheers >> Andy -- Sven Van Caekenberghe http://stfx.eu Smalltalk is the Red Pill

Jan. 23, 2013

Re: [Pharo-project] OrderedCollections and File I/O

by Edgar J. De Cleene

On 1/23/13 6:25 AM, "Marcus Denker" <marcus.denker(a)inria.fr> wrote: > Keep in mind that ReferenceStream is deprecated in 2.0 > > The nice thing about Fuel is: > > -> Very much faster > -> output more compact > -> you can store *any* kind of objects. > (even the one you are sure you can't, like the Context (stack > frame)) > > Marcus Yes , I know Fuel is a great advance, but he ask about simple thing. And also I said Pharo 1.4 and not Pharo 2.0. Thanks to Mariano and others , wish Fuel become Squeak citizen soon, for Cuis guess should wait longer. Edgar

Jan. 23, 2013

Re: [Pharo-project] OrderedCollections and File I/O

by Nicolas Cellier

+1 in the context of Pharo, no doubt, Fuel is THE solution Squeak has to support legacy ReferenceStream a bit longer, but even in Squeak I would inquire about Fuel 2013/1/23 Marcus Denker <marcus.denker(a)inria.fr>: > > On Jan 23, 2013, at 10:21 AM, "Edgar J. De Cleene" <edgardec2005(a)gmail.com> wrote: > >>> >> >> File in the attached in Pharo 1.4, Squeak or Cuis and you have a way of >> save / load simple Objects into the three forks. >> >> And if you speak Spanish, we have >> http://ar.groups.yahoo.com/group/squeakRos/ >> >> Join us and ask any question. >> >> Edgar >> >> <ObjectCompatibleENH.1.cs> > > Keep in mind that ReferenceStream is deprecated in 2.0 > > The nice thing about Fuel is: > > -> Very much faster > -> output more compact > -> you can store *any* kind of objects. > (even the one you are sure you can't, like the Context (stack frame)) > > Marcus > >

Jan. 23, 2013