Hi,

2014-09-15 10:28 GMT+02:00 Tudor Girba <tudor@tudorgirba.com>:

Hi,

On Mon, Sep 15, 2014 at 9:57 AM, Thierry Goubier <thierry.goubier@gmail.com> wrote:

2014-09-15 6:44 GMT+02:00 Tudor Girba <tudor@tudorgirba.com>:
Hi,

As you could see in the Pillar highlighting support, Rubric can now handle any syntax highlighting in a rather straightforward way. You can also easily embed this in a Glamour browser.

I'll have a look. I find that PluggableTextMorph automatic behavior (smart characters, styling) is very nice when doing Smalltalk code, but a pain in anything else. I'd like to see that Smalltalk-customized code be refactored elsewhere than in the editor.

What I'd like to see is the ability to customize the behavior for islands of text inside a text (recursively of course :)).

This is not yet possible. I would hope we get this to work within the context of TxText.

I was thinking of having a look at TxText ;)

I noticed that I have already started to reimplement part of the StringHolder / PluggableTextMorph logic outside it (via menu command objects / keymapping shortcuts instead of using the hardcoded versions inside the Morph or Model object) and I wanted to see if this was already done (or prepared for) in TxText.

I think that, even in the context of the current TextEditor, I could customize based on islands based on what I already have. But I have to spend some time on it.

�

�

You can see the example in the GT-InspectorExtensions-Pillar. You can load the code in a Moose image like this:
Gofer new
� � � � smalltalkhubUser: 'Pier' project: 'Pillar';
� � � � configuration;
� � � � loadStable.
Gofer new
� � � �smalltalkhubUser: 'JanKurs' project: 'PetitParser';
� � � �configurationOf: #PetitParserIslands;
� � � �load.
#ConfigurationOfPetitParserIslands asClass loadDevelopment.
Gofer new
� � � �smalltalkhubUser: 'Moose' project: 'GToolkit';
� � � �package: 'GT-InspectorExtensions-Pillar';
� � � �load.

Ok. The Moose image is a bit of a down to me when trying to understand code because code exploration capabilities are limited / non obvious compared to what I use in Pharo.

What do you mean? What is different for you in this area?

�

I'm accustomed to the smart suggestion stuff as I implemented it, makes it very easy to drill down in code...

What I use a lot:

Right mouse button on a global, select 'Browse' : inspect the class variable or open a browser on the class (depending on what it is). Also valid in a class #subclass: display in the browser, to browse the super class.

Implementors / sendors of / accesses to instance var under right mouse button: open a full browser showing all the resulting methods (and not a message list :( ). Also works in the string containing instance variable names in a class definition pane.

Double click on a package / class -> open a browser scoped to that package / class (where all search operations are limited to that scope).

Ctrl+F -> open the finder toolbar on top of the browser, with reduced gui (no package choice) : result of the finder will appear in a new browser, and finder is scoped by the initial browser scope.

So, when I end up in Nautilus, I just find myself very clumsy looking around (which is not good for my self-esteem ;) ) Not forgetting the fact that opening a Nautilus browser takes a lot longer. In Moose also, GTInspector is slow.

Take it as a different experience on the GUI side; I know I invest a lot in the motor skill part of the GUI, and this has an effect when I switch platform.

Thierry

�