March 2014 - Pharo-dev - lists.pharo.org

Re: [Pharo-dev] Crazy Smalltalk code snippets

by Frank Shearar

On 24 March 2014 23:33, phil(a)highoctane.be <phil(a)highoctane.be> wrote: > On Tue, Mar 25, 2014 at 12:11 AM, Eliot Miranda <eliot.miranda(a)gmail.com> > wrote: >> >> >> >> >> On Mon, Mar 24, 2014 at 4:02 PM, phil(a)highoctane.be <phil(a)highoctane.be> >> wrote: >>> >>> I am curious. >>> >>> Maybe on the first one, substitute something when a DNU is encountered. >>> Like logging undefined receivers. >>> >>> I wonder how the next one behaves on JITted methods. I fear that offset >>> errors may lead to weird errors. >> >> >> It *should* "just work" :-) (provided the jump distance of -12 is >> correct). The VM traps assignments to variables of contexts, and converts >> them to vanilla contexts. Therefore the jump doesn't occur in JITTED code >> but back in normal interpreted bytecode. >> > > How magical! > >>> >>> On Mon, Mar 24, 2014 at 11:51 PM, Pavel Krivanek >>> <pavel.krivanek(a)gmail.com> wrote: >>>> >>>> Who can find the most useful usage of this? >>>> >>>> thisContext instVarNamed: #receiver put: 42. >>>> self factorial >>>> >>>> GOTO statement in Pharo: >>>> >>>> FileStream stdout nextPutAll: 'Hello world'; lf. >>>> thisContext jump: -12. >>>> Let's collect the next ones :-) >> >> >> this is my favourite, and would lock-up a strict blue book VM. (I once >> locked up Allen Wirfs-Brock's 4404 with this and he wasn't best pleased >> [forgive me Allen]) >> >> >> | a | >> a := Array with: #perform:withArguments: with: nil. >> a at: 2 put: a. >> a perform: a first withArguments: a >> >> ;-) >> > > > The infinite is near. > > Not as blocking, but still... (Alt-. interrupts this one) > > Crazy>>run > Continuation currentDo: [ :cc | here := cc ]. > here value: true. > > Crazy new run > > BTW, is there any support for partial continuations in Pharo? Continuations > like this one look like full continuations and that's *huge*. Yes: http://ss3.gemstone.com/ss/Control.html Last time I checked I did need to add a shim (see the ControlPharo package), but it did load cleanly and pass all its own tests. Control not only provides a convenient way of making partial continuations (of the shift/reset sort, if you're familiar with the literature), but also _delimited_ dynamic variables. As soon as you start stack-slicing with partial continuations, you quickly find that standard implementations of dynamic variables fail in all sorts of nasty ways. Fundamentally, "normal" dynamic variables _cannot_ work cleanly with partial continuations because they either close over too much of the dynamic environment, or too little. Here's some reading on the topic: [1] http://okmij.org/ftp/Computation/dynamic-binding.html [2] http://www.cs.rutgers.edu/~ccshan/dynscope/DDBinding.pdf [3] http://www.lshift.net/blog/2012/06/27/resumable-exceptions-can-macro-expres… frank >>>> Cheers, >>>> -- Pavel >>> >>> >> >> >> >> -- >> best, >> Eliot > >

March 25, 2014

Re: [Pharo-dev] Using Spec based UI in Nautilus Plugins

by Benjamin

MethodToolbar new buildWithSpec Ben On 25 Mar 2014, at 10:09, phil(a)highoctane.be wrote: > This doesn't work for building the MathodToolbar widget. > > MethodToolbar defaultSpec buildWithSpec > > #DNU SpecLayout>>buildWithSpec > > MethodToolbar class>>defaultSpec > <spec> > ^ SpecLayout composed > newRow: [ :row | > row > add: #(model browseModel); > add: #(model sendersModel); > add: #(model implementorsModel); > add: #(model versionModel); > add: #(model dropListModel) ]; > yourself > > > > What to do? > > Phil > > On Tue, Mar 25, 2014 at 9:51 AM, Benjamin <benjamin.vanryseghem.pharo(a)gmail.com> wrote: > You create your own subclass of AbstractNautilusPlugin in where the display method will look like > > display > ^ mySpec buildWithSpec > > This should do the trick > > Ben > > On 25 Mar 2014, at 08:41, phil(a)highoctane.be wrote: > >> I wanted to use a MethodToolbar as part of a Nautilus plugin. >> >> Ok, I can use the widget in with an openWithSpec but how can I put this into a widget that Nautilus will accept as a plugin? >> >> >> | b | >> b := MethodToolbar new. >> b openWithSpec. >> b method: <method from plugin...> >> >> >> Phil >> > >

March 25, 2014

Re: [Pharo-dev] [Pharo-users] installing gtoolkit

by Tudor Girba

Hi, On Tue, Mar 25, 2014 at 10:02 AM, phil(a)highoctane.be <phil(a)highoctane.be>wrote: > On Tue, Mar 25, 2014 at 8:17 AM, Tudor Girba <tudor(a)tudorgirba.com> wrote: > >> Hi, >> >> If you want to install GToolkit (the project that contains GTInspector, >> GTPlayground and GTDebugger) in your image, you can follow one of the >> options as described here: >> http://www.humane-assessment.com/blog/installing-gtoolkit >> >> If you give it a try, please let us know how it works for you. >> > > Looks *very* nice. Takes a while to get used to. > Happy. > Now, the gtoolkit image I used didn't respected the Pharo3 keybindings > that allow one to open a workspace (Alt-o Alt-w / Cmd-o Cmd-w). > It would be nice to have them so that opening the Playground is done in a > snap. > The global shortcuts are disabled because they conflict with the Cmd+o shortcuts used everywhere in the GTInspector and GTPlayground. We have an issue open on this, but we did not yet come to a conclusion. It would be useful to have more eyeballs looking at it: https://pharo.fogbugz.com/f/cases/13102/Sequences-of-bindings-hide-individu… > Question: if one has opened a Miller column too far, how can we close that > one? It was annoying to have to use to precisely use the scrollbar down > there to get back to where I was. And if the column is not out of view, it > keeps on updating and moves the thing to the right. Kind of ping-pong style. > You cannot :(. The current implementation of the Miller column widget is not at all optimal. I have a new design that does not rely on a scrollbar at all, but to get that to work, we would need a bit more effort involved in building the necessary widgets. It would be great to have more contributions there. If anyone is interested, please let me know. > I'll load it in my current development project and see how it goes (that > superwhite theme of Moose is just an eye killer for me). > Great. Let me know how it works and what quirks you get into. Cheers, Doru > Phil > > >> >> Cheers, >> Doru >> >> -- >> www.tudorgirba.com >> >> "Every thing has its own flow" >> > > -- www.tudorgirba.com "Every thing has its own flow"

March 25, 2014

Re: [Pharo-dev] [Pharo-users] installing gtoolkit

by Sven Van Caekenberghe

On 25 Mar 2014, at 10:02, phil(a)highoctane.be wrote: > Question: if one has opened a Miller column too far, how can we close that one? It was annoying to have to use to precisely use the scrollbar down there to get back to where I was. And if the column is not out of view, it keeps on updating and moves the thing to the right. Kind of ping-pong style. I have had a similar remark in the past: I want to be able to close the extra columns. It all takes so much space and it a distraction once you no longer need them.

March 25, 2014

Re: [Pharo-dev] Using Spec based UI in Nautilus Plugins

by phil＠highoctane.be

This doesn't work for building the MathodToolbar widget. MethodToolbar defaultSpec buildWithSpec #DNU SpecLayout>>buildWithSpec MethodToolbar class>>defaultSpec <spec> ^ SpecLayout composed newRow: [ :row | row add: #(model browseModel); add: #(model sendersModel); add: #(model implementorsModel); add: #(model versionModel); add: #(model dropListModel) ]; yourself What to do? Phil On Tue, Mar 25, 2014 at 9:51 AM, Benjamin < benjamin.vanryseghem.pharo(a)gmail.com> wrote: > You create your own subclass of AbstractNautilusPlugin in where the > display method will look like > > display > ^ mySpec buildWithSpec > > This should do the trick > > Ben > > On 25 Mar 2014, at 08:41, phil(a)highoctane.be wrote: > > I wanted to use a MethodToolbar as part of a Nautilus plugin. > > Ok, I can use the widget in with an openWithSpec but how can I put this > into a widget that Nautilus will accept as a plugin? > > > | b | > b := MethodToolbar new. > b openWithSpec. > b method: <method from plugin...> > > > Phil > > >

March 25, 2014

Re: [Pharo-dev] installing gtoolkit

by phil＠highoctane.be

On Tue, Mar 25, 2014 at 8:17 AM, Tudor Girba <tudor(a)tudorgirba.com> wrote: > Hi, > > If you want to install GToolkit (the project that contains GTInspector, > GTPlayground and GTDebugger) in your image, you can follow one of the > options as described here: > http://www.humane-assessment.com/blog/installing-gtoolkit > > If you give it a try, please let us know how it works for you. > Looks *very* nice. Takes a while to get used to. Now, the gtoolkit image I used didn't respected the Pharo3 keybindings that allow one to open a workspace (Alt-o Alt-w / Cmd-o Cmd-w). It would be nice to have them so that opening the Playground is done in a snap. Question: if one has opened a Miller column too far, how can we close that one? It was annoying to have to use to precisely use the scrollbar down there to get back to where I was. And if the column is not out of view, it keeps on updating and moves the thing to the right. Kind of ping-pong style. I'll load it in my current development project and see how it goes (that superwhite theme of Moose is just an eye killer for me). Phil > > Cheers, > Doru > > -- > www.tudorgirba.com > > "Every thing has its own flow" >

March 25, 2014

Re: [Pharo-dev] Using Spec based UI in Nautilus Plugins

by Benjamin

You create your own subclass of AbstractNautilusPlugin in where the display method will look like display ^ mySpec buildWithSpec This should do the trick Ben On 25 Mar 2014, at 08:41, phil(a)highoctane.be wrote: > I wanted to use a MethodToolbar as part of a Nautilus plugin. > > Ok, I can use the widget in with an openWithSpec but how can I put this into a widget that Nautilus will accept as a plugin? > > > | b | > b := MethodToolbar new. > b openWithSpec. > b method: <method from plugin...> > > > Phil >

March 25, 2014

Re: [Pharo-dev] Crazy Smalltalk code snippets

by phil＠highoctane.be

On Tue, Mar 25, 2014 at 8:36 AM, Norbert Hartl <norbert(a)hartl.name> wrote: > > > Am 25.03.2014 um 00:33 schrieb "phil(a)highoctane.be" <phil(a)highoctane.be>: > > On Tue, Mar 25, 2014 at 12:11 AM, Eliot Miranda <eliot.miranda(a)gmail.com>wrote: > >> >> >> >> On Mon, Mar 24, 2014 at 4:02 PM, phil(a)highoctane.be <phil(a)highoctane.be>wrote: >> >>> I am curious. >>> >>> Maybe on the first one, substitute something when a DNU is encountered. >>> Like logging undefined receivers. >>> >>> I wonder how the next one behaves on JITted methods. I fear that offset >>> errors may lead to weird errors. >>> >> >> It *should* "just work" :-) (provided the jump distance of -12 is >> correct). The VM traps assignments to variables of contexts, and converts >> them to vanilla contexts. Therefore the jump doesn't occur in JITTED code >> but back in normal interpreted bytecode. >> >> > How magical! > > >> On Mon, Mar 24, 2014 at 11:51 PM, Pavel Krivanek < >>> pavel.krivanek(a)gmail.com> wrote: >>> >>>> Who can find the most useful usage of this? >>>> >>>> thisContext instVarNamed: #receiver put: 42. >>>> self factorial >>>> >>>> GOTO statement in Pharo: >>>> >>>> FileStream stdout nextPutAll: 'Hello world'; lf. >>>> thisContext jump: -12. >>>> Let's collect the next ones :-) >>>> >>> >> this is my favourite, and would lock-up a strict blue book VM. (I once >> locked up Allen Wirfs-Brock's 4404 with this and he wasn't best pleased >> [forgive me Allen]) >> >> >> | a | >> a := Array with: #perform:withArguments: with: nil. >> a at: 2 put: a. >> a perform: a first withArguments: a >> >> ;-) >> >> > > The infinite is near. > > Not as blocking, but still... (Alt-. interrupts this one) > > Crazy>>run > Continuation currentDo: [ :cc | here := cc ]. > here value: true. > > Crazy new run > > BTW, is there any support for partial continuations in Pharo? > Continuations like this one look like full continuations and that's *huge*. > > If you mean delimited continuations have a look here: > > > http://www.lshift.net/blog/2011/04/20/direct-implementation-of-shiftreset-i… > > Thanks. I saw a mention of this but not that article. Now, I loaded the code in Pharo3 and there are missing methods in BlockContext... ControlTests>>testNestedShiftResetNest (I guess other things would say the same) is broken testNestedShiftResetsNest | innerCont outerCont | [1 + [ :k | outerCont := k. "stack reified as [ :x | 1 + x ]" k value: [ 2 + [ :kk | innerCont := kk. "stack reified as [ :x | 2 + x ]" kk value: 0 ] shift ] reset ] shift ] reset. self assert: 3 equals: (innerCont value: 1). self assert: 2 equals: (outerCont value: 1). self assert: 5 equals: ((outerCont compose: [:x | x * 2]) value: 2) due to MethodContext's DNU: copyTo:bottomContextDo: [image: Inline image 1] There is a ContextPart: aContext blocks: dict but but looking like the same thing. Composition tests do work. Gurus welcome... If we put continuations in the Pharo image, why not put this as well? Phil > Norbert > > >>>> Cheers, >>>> -- Pavel >>>> >>> >>> >> >> >> -- >> best, >> Eliot >> > >

March 25, 2014

Re: [Pharo-dev] Trait1 Trait2 Trait3 in Smalltalk globals allClassesAndTraits

by Camille Teruel

On 24 mars 2014, at 23:30, phil(a)highoctane.be wrote: > On a fresh 3.0, there are Trait1, Trait2 and Trait3 in Smalltalk globals allClassesAndTraits > > These 3 look like they are equivalents to Trait. What are these for? That's just some test fixtures. > > Phil > >

March 25, 2014

Re: [Pharo-dev] Crazy Smalltalk code snippets

by Norbert Hartl

Am 24.03.2014 um 23:51 schrieb Pavel Krivanek <pavel.krivanek(a)gmail.com>: > Who can find the most useful usage of this? > > thisContext instVarNamed: #receiver put: 42. > self factorial > > GOTO statement in Pharo: > > FileStream stdout nextPutAll: 'Hello world'; lf. > thisContext jump: -12. > > Let's collect the next ones :-) > | injectedContext returnValue | injectedContext := [ returnValue := self dummySendToSkip. block value. returnValue ] asContext. injectedContext pc: injectedContext startpc + 2. injectedContext swapSender: (self targetContext swapSender: injectedContext) I used this once. It is hackish but I still like it because I donât know how to solver that better :) Norbert

March 25, 2014