- Pharo-dev - lists.pharo.org

Re: [Pharo-project] Vim Keys?

by Igor Stasenko

click on the top-right arrow in browser window and find "shortcuts description" menu item. On 30 November 2012 13:29, Sebastian Nozzi <sebnozzi(a)gmail.com> wrote: > Thanks for the whole information! > Actually I just wanted to know the key-combination (I expressed myself > wrong), but will keep this info for future reference. > > 2012/11/30 Benjamin <benjamin.vanryseghem.pharo(a)gmail.com>: >> >> On Nov 30, 2012, at 10:25 AM, Goubier Thierry wrote: >> >>> Le 30/11/2012 09:57, Sebastian Nozzi a Ã©crit : >>>> When we are at it... how do shortcuts in Pharo 2 currently work? >>> >>> Ouch. I can describe what I know about key event processing, and maybe you will understand what's happening. It will be a good exercise for me to see if I got that stuff right. >>> >>> From first to process to last to process, once the key event is generated and given to the Morph which has the focus: >>> >>> 1 - The morph keymapping dispatch. >>> This one is multilevel in nature. Here, so it goes (simplified, I'm not listing where platform differences are taken in account). The key event is matched against shortcuts defined in keymaps (and a partial match is possible if it is a multi-key shortcut). >>> 1.1 - Direct keymapping : shortcuts added by on: do: to the KMDispatcher of the Morph instance. >>> 1.2 - Named keymaps. Keymaps defined elsewhere and attached to that morph keymap dispatcher. >>> 1.3 - Global named keymaps. Keymaps associated with the morph Class or one of it's superclass (i.e. a Morph class global keymap will apply to all morphs). >>> 1.4 - If no match, go to 1.1 with the owner of the morph and repeat. Do that until you reach the World (Pharo top-level window). There, if keymapping hasn't matched, go to 2. >>> >>> 2 - The morph keyStroke: handling. Normal keys, navigation keys, hardcoded shortcuts (TextMorph for example). >>> -- In some cases (some! No, often :(!) keystrokes are sent to other objects or Morphs: navigation, shortcuts, etc... >>> >>> 3 - The morph eventHandler : here a model can trap any key event or shortcut. >>> >>> If a match happen in any of those, the key event is usually said to be processed and we go out of the processing loop (no more matches) >>> >>> So, for a given shortcut being processed, it may be hard to find where it has been caught. Only the Keymapping dispatch has a debugging feature (with KMLog). >>> >>>> In Pharo 1.4, in the class browser, I used to hit Ctrl-F to open "Find >>>> Class", but it has no effect in Pharo 2. >>> >>> This one is easier. In Nautilus, there is a shortcut browser which lists all the defined shortcuts. I believe that the find class is a multi-key shortcut. >> >> Indeed , cmd+f,c >> >> Ben >> :) >> >>> >>>> 2012/11/30 Goubier Thierry <thierry.goubier(a)cea.fr>: >>>>> Le 29/11/2012 20:52, âking a Ã©crit : >>>>> >>>>>> Hi all. I'm extremely new to Smalltalk, but I was wondering if there was >>>>>> a way to make Pharo have vi-keys (or even better, vim-keys)? >>>>>> >>>>>> Thanks! >>>>>> ââ >>>>> >>>>> >>>>> It's underway. The infrastructure is moving to a better (unified) way of >>>>> specifying and handling shortcuts, and then vi and vim-keys and emacs and >>>>> others should become available. >>>>> >>>>> Bad point: I should help with that effort, but given my deadlines for the >>>>> end of the year, not sure I'll be of much help :(. >>>>> >>>>> Thierry >>>>> -- >>>>> Thierry Goubier >>>>> CEA list >>>>> Laboratoire des Fondations des SystÃ¨mes Temps RÃ©el EmbarquÃ©s >>>>> 91191 Gif sur Yvette Cedex >>>>> France >>>>> Phone/Fax: +33 (0) 1 69 08 32 92 / 83 95 >>>>> >>>> >>>> >>>> >>> >>> >>> -- >>> Thierry Goubier >>> CEA list >>> Laboratoire des Fondations des SystÃ¨mes Temps RÃ©el EmbarquÃ©s >>> 91191 Gif sur Yvette Cedex >>> France >>> Phone/Fax: +33 (0) 1 69 08 32 92 / 83 95 >>> >> >> > -- Best regards, Igor Stasenko.

Nov. 30, 2012

Re: [Pharo-project] Vim Keys?

by StÃ©phane Ducasse

On Nov 30, 2012, at 9:57 AM, Sebastian Nozzi wrote: > When we are at it... how do shortcuts in Pharo 2 currently work? > > In Pharo 1.4, in the class browser, I used to hit Ctrl-F to open "Find > Class", but it has no effect in Pharo 2. for now but we are discussing it :) ctrl-f + ctrl c => find class > > 2012/11/30 Goubier Thierry <thierry.goubier(a)cea.fr>: >> Le 29/11/2012 20:52, âking a Ã©crit : >> >>> Hi all. I'm extremely new to Smalltalk, but I was wondering if there was >>> a way to make Pharo have vi-keys (or even better, vim-keys)? >>> >>> Thanks! >>> ââ >> >> >> It's underway. The infrastructure is moving to a better (unified) way of >> specifying and handling shortcuts, and then vi and vim-keys and emacs and >> others should become available. >> >> Bad point: I should help with that effort, but given my deadlines for the >> end of the year, not sure I'll be of much help :(. >> >> Thierry >> -- >> Thierry Goubier >> CEA list >> Laboratoire des Fondations des SystÃ¨mes Temps RÃ©el EmbarquÃ©s >> 91191 Gif sur Yvette Cedex >> France >> Phone/Fax: +33 (0) 1 69 08 32 92 / 83 95 >> >

Nov. 30, 2012

Re: [Pharo-project] Versioning external packages

by StÃ©phane Ducasse

read the monticello chapter in the book :) On Nov 27, 2012, at 12:12 PM, roberto.minelli(a)usi.ch wrote: > Hi all, > > I'm using Monticello to version my source code. > > Suppose this use case: I need to add a message to the OrderedCollection class and I want my message to be versioned within my repository (i.e., when an user downloads my package she also get the new message in the OrderedCollection class. > > And another one, suppose I create a new Layout for Roassal that suits my needs. I will create a class in the category 'Roassal-Layout' which is not versioned in my repository. Is there a way to version it? > > Cheers, > Roberto

Nov. 30, 2012

Re: [Pharo-project] Changes in announcements

by StÃ©phane Ducasse

thanks ezequiel. Stef On Nov 27, 2012, at 11:09 AM, Ezequiel Lamonica wrote: > Hi, > > > > > I was reviewing with Esteban about the lack of some announcements in the system and some bugs in other ones. > > I've been working on that for this week. > > This is the list: > > -MethodRename > -Snapshot > -Session (i don't know if this is necessary) > -VariableAdded (one for every kind) > -VariableRemoved (one for every kind) > -Protocols (add/modif/remove/rename) > -TraitComposition > - When a package is created, the announcement is not released > This only happens when a class is created with that category. > -When I modify a class and also I change its category, then only class recategorized announcement is sent > -Package removed is not sent

Nov. 30, 2012

Re: [Pharo-project] Alt-. does not work anymore on Linux or?

by StÃ©phane Ducasse

We cleaned all the evt at: 1. But did not find the time to push it in the system. We will work on that. On Nov 29, 2012, at 2:33 PM, Torsten Bergmann wrote: >> I can try. What I find a bit "odd" is that when I press Alt-. it >> actually prints a $. (say in a Workspace) just as if I had *only* >> pressed dot. That may be the main problem here - Alt is not properly >> detected? > > Yes same here. But some Transcript outputs into UserInterruptHandler>>handleEvent: > > and you will notice that it comes with different "(evt at: 4)". > > Yes - to me it also looks as if ALT is not properly detected > ... > > > The default interrupt key value in UserInterruptHandler(class)>>new > is > > ($. asciiValue bitOr: 16r0800) > > So I guess this bit enabeling is for ALT (CMD). > > Dont know why we dont use a shared pool with proper constants > > KEY_CMD_BIT := 16r0800 > > > ($. asciiValue bitOr: KEY_CMD_BIT) > > which would be much clearer. At least to me. > > > The code in UserInterruptHandler>>handleEvent: could profit > from constants as well, I hate these "(evt at: 4)" and > "(evt at: 3)" and "(evt at: 5)" with bit manipulation using > magic 8 ... > > >

Nov. 30, 2012

Re: [Pharo-project] Alt-. does not work anymore on Linux or?

by StÃ©phane Ducasse

HI goran can you open a bug entry? Tx! Stef On Nov 29, 2012, at 12:28 PM, GÃ¶ran Krampe wrote: > Hey! > > I just realized that alt-. doesn't seem to fly on my Ubuntu 12.10 in Pharo 1.4 nor 2.0 one-clicks. > > It *does* however work fine in Squeak 4.3 one-click. > > In Squeak I can: > Run "(Delay forSeconds: 5) wait" and interrupt it with alt-. > Or just press Alt-. and get a debugger. > > Also it seems that Process browser in Squeak is opened up *without* having any process be suspended. Not sure why Pharo is different there and shows the UI process as suspended? > > regards, GÃ¶ran > >

Nov. 30, 2012

Re: [Pharo-project] [ann] glamorous debugger (alpha)

by Tudor Girba

Thanks :). This should not be difficult, and it's a nice idea. Next week, Andrei will be visiting Lille, so perhaps this would be a good opportunity to play with things :). Cheers, Doru On Fri, Nov 30, 2012 at 12:36 PM, Damien Cassou <damien.cassou(a)gmail.com>wrote: > Hi Doru, > > that's great news and another sign that Glamour is very powerful. How hard > would it be to show the values for the current expression in the list of > variables? For example, while debugging the line > > self doSomething: (value at: 1) with: aStream next > > I would see in the list of variables the values for the expressions: > > value at: 1 > aStream next > > > > On Sun, Nov 25, 2012 at 9:42 AM, StÃ©phane Ducasse < > stephane.ducasse(a)inria.fr> wrote: > >> No so far only for 1.4 >> >> Stef >> >> On Nov 25, 2012, at 1:48 AM, Camillo Bruni wrote: >> >> >> >> >> If you are interested in playing with it, you can either find it in >> the Moose image, or you can load it in Pharo 1.4 via: >> >> Gofer new >> >> squeaksource: 'glamoroust'; >> >> package: 'ConfigurationOfGlamoroust'; >> >> load. >> >> (Smalltalk at: #ConfigurationOfGlamoroust) loadDevelopment >> > >> > is it supposed to load under 2.0? I get an issue with >> ConfigurationOfRPackage.. >> >> >> > > > -- > Damien Cassou > http://damiencassou.seasidehosting.st > > "Success is the ability to go from one failure to another without losing > enthusiasm." > Winston Churchill > -- www.tudorgirba.com "Every thing has its own flow"

Nov. 30, 2012

Re: [Pharo-project] Indent code

by Max Leske

On 30.11.2012, at 13:28, Sebastian Nozzi <sebnozzi(a)gmail.com> wrote: >>> When I select a few lines of text in Pharo 2.0 code >>> editor and hit TAB it is not indented as in >>> other IDE's. Instead it is deleted similar to >>> when one hits space key. >>> >>> Usually TAB in IDE code editors is used to >>> indent a selected block of text and SHIFT+TAB >>> is used for the reverse (outdent). > > I also expected that. Would be cool if this was the default behavior. > >> on mac: ctrl+alt+r / l > > This makes sense from a language perspective (right/left), but note > that not only you have to press more keys, but also is more tedious if > you are adjusting the indentation. With this I mean if you want to > quickly do left/right operations after another. > I didn't say it's good, just how it currently works :) +1 for better shortcuts (but checkout the currently active "Vim Keys?" thread for that) Max

Nov. 30, 2012

Re: [Pharo-project] Vim Keys?

by Sebastian Nozzi

Thanks for the whole information! Actually I just wanted to know the key-combination (I expressed myself wrong), but will keep this info for future reference. 2012/11/30 Benjamin <benjamin.vanryseghem.pharo(a)gmail.com>: > > On Nov 30, 2012, at 10:25 AM, Goubier Thierry wrote: > >> Le 30/11/2012 09:57, Sebastian Nozzi a Ã©crit : >>> When we are at it... how do shortcuts in Pharo 2 currently work? >> >> Ouch. I can describe what I know about key event processing, and maybe you will understand what's happening. It will be a good exercise for me to see if I got that stuff right. >> >> From first to process to last to process, once the key event is generated and given to the Morph which has the focus: >> >> 1 - The morph keymapping dispatch. >> This one is multilevel in nature. Here, so it goes (simplified, I'm not listing where platform differences are taken in account). The key event is matched against shortcuts defined in keymaps (and a partial match is possible if it is a multi-key shortcut). >> 1.1 - Direct keymapping : shortcuts added by on: do: to the KMDispatcher of the Morph instance. >> 1.2 - Named keymaps. Keymaps defined elsewhere and attached to that morph keymap dispatcher. >> 1.3 - Global named keymaps. Keymaps associated with the morph Class or one of it's superclass (i.e. a Morph class global keymap will apply to all morphs). >> 1.4 - If no match, go to 1.1 with the owner of the morph and repeat. Do that until you reach the World (Pharo top-level window). There, if keymapping hasn't matched, go to 2. >> >> 2 - The morph keyStroke: handling. Normal keys, navigation keys, hardcoded shortcuts (TextMorph for example). >> -- In some cases (some! No, often :(!) keystrokes are sent to other objects or Morphs: navigation, shortcuts, etc... >> >> 3 - The morph eventHandler : here a model can trap any key event or shortcut. >> >> If a match happen in any of those, the key event is usually said to be processed and we go out of the processing loop (no more matches) >> >> So, for a given shortcut being processed, it may be hard to find where it has been caught. Only the Keymapping dispatch has a debugging feature (with KMLog). >> >>> In Pharo 1.4, in the class browser, I used to hit Ctrl-F to open "Find >>> Class", but it has no effect in Pharo 2. >> >> This one is easier. In Nautilus, there is a shortcut browser which lists all the defined shortcuts. I believe that the find class is a multi-key shortcut. > > Indeed , cmd+f,c > > Ben > :) > >> >>> 2012/11/30 Goubier Thierry <thierry.goubier(a)cea.fr>: >>>> Le 29/11/2012 20:52, âking a Ã©crit : >>>> >>>>> Hi all. I'm extremely new to Smalltalk, but I was wondering if there was >>>>> a way to make Pharo have vi-keys (or even better, vim-keys)? >>>>> >>>>> Thanks! >>>>> ââ >>>> >>>> >>>> It's underway. The infrastructure is moving to a better (unified) way of >>>> specifying and handling shortcuts, and then vi and vim-keys and emacs and >>>> others should become available. >>>> >>>> Bad point: I should help with that effort, but given my deadlines for the >>>> end of the year, not sure I'll be of much help :(. >>>> >>>> Thierry >>>> -- >>>> Thierry Goubier >>>> CEA list >>>> Laboratoire des Fondations des SystÃ¨mes Temps RÃ©el EmbarquÃ©s >>>> 91191 Gif sur Yvette Cedex >>>> France >>>> Phone/Fax: +33 (0) 1 69 08 32 92 / 83 95 >>>> >>> >>> >>> >> >> >> -- >> Thierry Goubier >> CEA list >> Laboratoire des Fondations des SystÃ¨mes Temps RÃ©el EmbarquÃ©s >> 91191 Gif sur Yvette Cedex >> France >> Phone/Fax: +33 (0) 1 69 08 32 92 / 83 95 >> > >

Nov. 30, 2012

Re: [Pharo-project] Indent code

by Sebastian Nozzi

>> When I select a few lines of text in Pharo 2.0 code >> editor and hit TAB it is not indented as in >> other IDE's. Instead it is deleted similar to >> when one hits space key. >> >> Usually TAB in IDE code editors is used to >> indent a selected block of text and SHIFT+TAB >> is used for the reverse (outdent). I also expected that. Would be cool if this was the default behavior. > on mac: ctrl+alt+r / l This makes sense from a language perspective (right/left), but note that not only you have to press more keys, but also is more tedious if you are adjusting the indentation. With this I mean if you want to quickly do left/right operations after another.

Nov. 30, 2012