- Pharo-users - lists.pharo.org

Re: [Pharo-users] trait question

by Igor Stasenko

On 1 July 2013 14:01, Camille Teruel <camille.teruel(a)gmail.com> wrote: > > On 1 juil. 2013, at 12:19, Igor Stasenko wrote: > >> On 30 June 2013 15:48, Damien Cassou <damien.cassou(a)gmail.com> wrote: >>> On Sun, Jun 30, 2013 at 2:54 PM, Camillo Bruni <camillobruni(a)gmail.com> wrote: >>>> But you know that you can do this right now? >>> >>> >>> I think these examples only show that some verifications are missing. >>> I don't know any source code using that. To me it's more a bug than a >>> feature. I would keep things simple: applying a trait to a class >>> applies the classTrait to the metaclass. >>> >> but that's far less simpler than: >> - applying trait to behavior (object), applies it only to that object >> and nothing else (no meta-mess). > > Yet the current trait model is based on trait and class trait, and that's consistent with the class model. > If you want trait to be applied only on one side: > - we have to refactor a lot of stuff, because a lot of our current code rely on this model. > - you have to provide a new way for managing instance/class-side dependancies: > For example, if one the class side you have: "self class blah" how can you ensure that the class side has the implementation you expect? it is not responsibility of trait to reason whether "self class" gives right/wrong object to work with. #class is just a message sent to receiver.. you may think it is special, but there's nothing special in it comparing to any other message you sending to object(s). in any case, when developer applies trait on class side, he knows what he doing, isn't? > > On the other hand if you want a trait to be applied only on the instance side: provide only instance-side methods. > And if you want a trait to be applied only on the class-side: provides only class-side methods. > So everything's fine with this regard IMO. Don't change it, just ensure the missing verification in the class builder. > Fine? Ok, how to apply trait which has single method, to make this method appear on both class and instance sides? (Copy/pasting same method twice is not an answer). I am not asking for changing things. But i think that implementation would be simpler, if Trait do not have to support class/metaclass duality. -- Best regards, Igor Stasenko.

July 1, 2013

Re: [Pharo-users] Does Morphic include a control similar to this? (jpeg attached)

by Benjamin

I think such a behaviour is doable in Glamour, which means there is most probably a morph behind. Maybe Doru could answer better than me :) Ben On Jul 1, 2013, at 2:36 PM, basilmir <mircea(a)unom.ro> wrote: > Does Morphic include a control similar to this? (jpeg attached) > > I need a two column graphics element (key, value pair). Some of these key > value pairs are grouped under a category which can (optionally) collapse. Is > there something similar in Morphic - Pharo, Squeak? > > Thank you! > > <http://forum.world.st/file/n4696579/F%C4%83r%C4%83_titlu.jpg> > > > > -- > View this message in context: http://forum.world.st/Does-Morphic-include-a-control-similar-to-this-jpeg-a… > Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com. >

July 1, 2013

Does Morphic include a control similar to this? (jpeg attached)

by basilmir

Does Morphic include a control similar to this? (jpeg attached) I need a two column graphics element (key, value pair). Some of these key value pairs are grouped under a category which can (optionally) collapse. Is there something similar in Morphic - Pharo, Squeak? Thank you! <http://forum.world.st/file/n4696579/F%C4%83r%C4%83_titlu.jpg> -- View this message in context: http://forum.world.st/Does-Morphic-include-a-control-similar-to-this-jpeg-a… Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

July 1, 2013

Re: [Pharo-users] trait question

by Camille Teruel

On 1 juil. 2013, at 12:19, Igor Stasenko wrote: > On 30 June 2013 15:48, Damien Cassou <damien.cassou(a)gmail.com> wrote: >> On Sun, Jun 30, 2013 at 2:54 PM, Camillo Bruni <camillobruni(a)gmail.com> wrote: >>> But you know that you can do this right now? >> >> >> I think these examples only show that some verifications are missing. >> I don't know any source code using that. To me it's more a bug than a >> feature. I would keep things simple: applying a trait to a class >> applies the classTrait to the metaclass. >> > but that's far less simpler than: > - applying trait to behavior (object), applies it only to that object > and nothing else (no meta-mess). Yet the current trait model is based on trait and class trait, and that's consistent with the class model. If you want trait to be applied only on one side: - we have to refactor a lot of stuff, because a lot of our current code rely on this model. - you have to provide a new way for managing instance/class-side dependancies: For example, if one the class side you have: "self class blah" how can you ensure that the class side has the implementation you expect? On the other hand if you want a trait to be applied only on the instance side: provide only instance-side methods. And if you want a trait to be applied only on the class-side: provides only class-side methods. So everything's fine with this regard IMO. Don't change it, just ensure the missing verification in the class builder. > >> -- >> Damien Cassou >> http://damiencassou.seasidehosting.st >> >> "Success is the ability to go from one failure to another without >> losing enthusiasm." >> Winston Churchill >> > > > > -- > Best regards, > Igor Stasenko. >

July 1, 2013

Re: [Pharo-users] trait question

by Igor Stasenko

On 30 June 2013 15:48, Damien Cassou <damien.cassou(a)gmail.com> wrote: > On Sun, Jun 30, 2013 at 2:54 PM, Camillo Bruni <camillobruni(a)gmail.com> wrote: >> But you know that you can do this right now? > > > I think these examples only show that some verifications are missing. > I don't know any source code using that. To me it's more a bug than a > feature. I would keep things simple: applying a trait to a class > applies the classTrait to the metaclass. > but that's far less simpler than: - applying trait to behavior (object), applies it only to that object and nothing else (no meta-mess). > -- > Damien Cassou > http://damiencassou.seasidehosting.st > > "Success is the ability to go from one failure to another without > losing enthusiasm." > Winston Churchill > -- Best regards, Igor Stasenko.

July 1, 2013

Re: [Pharo-users] how to draw this Shape??

by ClÃ©ment Bera

I don't know how to do it with Polymorph. You can do it with Athens (by polishing this code) : surf := AthensCairoSurface extent: 100@100. surf drawDuring: [:can | surf clear. can setPaint: Color red. can drawShape: ( can createPath: [:path | path absolute; moveTo: 0@25; curveVia: 25@0 to: 50@ 25; curveVia: 75@0 to: 100@ 25; lineTo: 50@75; lineTo: 0@25 ]) ]. Display getCanvas drawImage: surf asForm at: 25@25 Athens is by default in Pharo 3. If you are on Pharo 2, there is a configuration to load it somewhere : Gofer new smalltalkhubUser: 'Pharo' project: 'Athens'; configuration; load. ConfigurationOfAthens loadStable. But I don't know if you have to specify a version for Pharo 2.0 or something like that. 2013/7/1 irfankhan1 <irfankhan(a)hotmail.ca> > Hello I am trying to draw a Valentine heart shape as shown below, Border > of > this shape should be dashed bordered. > <http://forum.world.st/file/n4696507/Lavender-sachet-heart-shape.jpg> . I > think this can be done by PolygonMorph class, no idea how . > > > > -- > View this message in context: > http://forum.world.st/how-to-draw-this-Shape-tp4696507.html > Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com. > > -- ClÃ©ment BÃ©ra Mate Virtual Machine Engineer BÃ¢timent B 40, avenue Halley 59650 *Villeneuve d'Ascq*

July 1, 2013

how to draw this Shape??

by irfankhan1

Hello I am trying to draw a Valentine heart shape as shown below, Border of this shape should be dashed bordered. <http://forum.world.st/file/n4696507/Lavender-sachet-heart-shape.jpg> . I think this can be done by PolygonMorph class, no idea how . -- View this message in context: http://forum.world.st/how-to-draw-this-Shape-tp4696507.html Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

July 1, 2013

Re: [Pharo-users] ZdcSocketStream dnu #position

by StÃ©phane Ducasse

https://pharo.fogbugz.com/default.asp?11088 On Jun 30, 2013, at 9:11 PM, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > Hi Again, Paul, > > It was a typo ;-) > I added two new unit test as well. > > Here are the relevant commits: > > === > Name: Zinc-HTTP-SvenVanCaekenberghe.372 > Author: SvenVanCaekenberghe > Time: 30 June 2013, 9:06:10.101049 pm > UUID: 98ea090f-27d4-498e-8111-95e954d010a9 > Ancestors: Zinc-HTTP-SvenVanCaekenberghe.371 > > Fixed a typo in ZnStringEntity>>#readFrom: (thanks Paul DeBruicker for reporting the issue) > === > Name: Zinc-Tests-SvenVanCaekenberghe.191 > Author: SvenVanCaekenberghe > Time: 30 June 2013, 9:07:42.519249 pm > UUID: 25345906-5574-4fd3-a5d9-eeee711fdf58 > Ancestors: Zinc-Tests-SvenVanCaekenberghe.190 > > Added ZnClientTests>>testGet[Large]WideStringHTML > === > > You'll have to load #bleedingEdge to get these. > > Thanks again for testing the latest version and reporting the issue. > > Sven > > On 30 Jun 2013, at 20:19, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > >> Hi Paul, >> >> On 30 Jun 2013, at 19:38, Paul DeBruicker <pdebruic(a)gmail.com> wrote: >> >>> Hi - >>> >>> I've got both Zinc and Zodiac loaded into a Pharo 2 image. When I do a >>> ZnClient #get: to an http connection I'm getting the dnu mentioned in >>> the subject of this email. >>> >>> >>> The message send that causes it is in ZnStringEntity>>#readFrom: >>> >>> >>> ZnNetworkingUtils default socketStreamClass prints: 'ZdcSocketStream' >>> >>> Changing that to 'SocketStream' like so: >>> >>> ZnNetworkingUtils default socketStreamClass: SocketStream >>> >>> makes everything work . >>> >>> >>> Just wanted to let you know >>> >>> >>> Paul >> >> Thanks a lot for reporting this issue. >> >> It is a bit weird though: I normally don't use #position on principle, but apparently I did in some recent optimisation code, I also vaguely remember adding #position somewhere.. >> >> Furthermore, all unit tests are green on both 2.0 and latest 3.0 with ZdcSocketStream as default. >> >> Step one would be to add a unit test. >> >> Is it possible for you to give me the URL or a similar one exposing the problem - it would have to be related to non-ASCII, non-Latin-1 characters I guess. >> >> Thanks, >> >> Sven >> >> >> -- >> Sven Van Caekenberghe >> Proudly supporting Pharo >> http://pharo.org >> http://association.pharo.org >> http://consortium.pharo.org >> >> >> >> >> > >

June 30, 2013

Re: [Pharo-users] trait question

by Tudor Girba

Hi, Theoretically, I could do the same for inheritance as well. But, that would be too complex for most cases, so it is the implicit behavior. To me, the decision for trait handling is similar. Cheers, Doru On Sun, Jun 30, 2013 at 2:54 PM, Camillo Bruni <camillobruni(a)gmail.com>wrote: > On 2013-06-30, at 11:16, Sebastian Tleye <stleye(a)gmail.com> wrote: > > I also like this uniform because it's easier to program, > > Up to now i haven't had the need to separate them. > > > > 2013/6/30 Tudor Girba <tudor(a)tudorgirba.com> > >> +1 > >> > >> Keeping things uniform is a plus in this case, even though we might lose > >> some flexibility. > >> > >> Doru > > > But you know that you can do this right now? > > Object subclass: #Foo > uses: TMyTrait classTrait > > and the same on the class-side :/ > > Foo class > uses: TMyOtherTrait > instanceVariableNames: 'a b c' > > so from that point of view the current system allows for a messy/flexible > combination > of instance-side and class-side traits. > > > -- www.tudorgirba.com "Every thing has its own flow"

June 30, 2013

Re: [Pharo-users] ZdcSocketStream dnu #position

by Sven Van Caekenberghe

Hi Again, Paul, It was a typo ;-) I added two new unit test as well. Here are the relevant commits: === Name: Zinc-HTTP-SvenVanCaekenberghe.372 Author: SvenVanCaekenberghe Time: 30 June 2013, 9:06:10.101049 pm UUID: 98ea090f-27d4-498e-8111-95e954d010a9 Ancestors: Zinc-HTTP-SvenVanCaekenberghe.371 Fixed a typo in ZnStringEntity>>#readFrom: (thanks Paul DeBruicker for reporting the issue) === Name: Zinc-Tests-SvenVanCaekenberghe.191 Author: SvenVanCaekenberghe Time: 30 June 2013, 9:07:42.519249 pm UUID: 25345906-5574-4fd3-a5d9-eeee711fdf58 Ancestors: Zinc-Tests-SvenVanCaekenberghe.190 Added ZnClientTests>>testGet[Large]WideStringHTML === You'll have to load #bleedingEdge to get these. Thanks again for testing the latest version and reporting the issue. Sven On 30 Jun 2013, at 20:19, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > Hi Paul, > > On 30 Jun 2013, at 19:38, Paul DeBruicker <pdebruic(a)gmail.com> wrote: > >> Hi - >> >> I've got both Zinc and Zodiac loaded into a Pharo 2 image. When I do a >> ZnClient #get: to an http connection I'm getting the dnu mentioned in >> the subject of this email. >> >> >> The message send that causes it is in ZnStringEntity>>#readFrom: >> >> >> ZnNetworkingUtils default socketStreamClass prints: 'ZdcSocketStream' >> >> Changing that to 'SocketStream' like so: >> >> ZnNetworkingUtils default socketStreamClass: SocketStream >> >> makes everything work . >> >> >> Just wanted to let you know >> >> >> Paul > > Thanks a lot for reporting this issue. > > It is a bit weird though: I normally don't use #position on principle, but apparently I did in some recent optimisation code, I also vaguely remember adding #position somewhere.. > > Furthermore, all unit tests are green on both 2.0 and latest 3.0 with ZdcSocketStream as default. > > Step one would be to add a unit test. > > Is it possible for you to give me the URL or a similar one exposing the problem - it would have to be related to non-ASCII, non-Latin-1 characters I guess. > > Thanks, > > Sven > > > -- > Sven Van Caekenberghe > Proudly supporting Pharo > http://pharo.org > http://association.pharo.org > http://consortium.pharo.org > > > > >

June 30, 2013