C++ Stylistics

[Rory McGuire]

On Sun, 18 Jul 2010 23:21:14 +0200, BLS <windevguy at hotmail.de> wrote:

> On 18/07/2010 11:42, Justin Johansson wrote:
>> bearophile wrote:
>>> A Google talk in 2008 by Kevlin Henney, "C++ Stylistics":
>>> http://www.youtube.com/watch?v=zh8W4ZglOlw
>>>
>>> Some of the limits/problems of C++ he talks about are patched by Qt,
>>> like signals & slots, Q_FLAGS, Q_PROPERTY, Q_ENUMS:
>>> http://doc.trolltech.com/latest/moc.html
>>> GUIs have needs quite different from the features offered by standard
>>> C++: they don't need high speed processing, but they enjoy a good
>>> runtime reflection and few other things, most of them offered in some
>>> way by the Qt "Meta-Object Compiler".
>>>
>>> Bye,
>>> bearophile
>>
>> You may have missed it but I raised the topic of Qt, it's signals
>> and slots architecture and its meta object compiler, on this
>> newsgroup about two months ago under the subject
>>
>> "Can D be cute (Qt)"
>>
>> http://www.digitalmars.com/webnews/newsgroups.php?art_group=digitalmars.D&article_id=109832
>>
>>
>> Cheers
>> Justin Johansson
>>
>>
>
> I regret that this is such a silence discussion. In fact Lutger (sorry  
> don't know your full name) has implemented a fine QT like thread safe  
> Signal Slot library.
>
> // Q_FLAGS, Q_PROPERTY, Q_ENUMS and GUI
> Finally this means > we need more _traits() power.
>
> --We have, afaik.  no support for isProperty(), and to make  
> visual/interactive GUI development possible, isEvent() is also necessary.

Copied from http://digitalmars.com/d/2.0/phobos/std_traits.html
isProperty can be implemented using:

functionAttributes!(foo.func) & FunctionAttribute.PROPERTY





>
> -- Now that we know that Compile- to Run- time Reflection could be just  
> a mixin template away we have another good reason to ask for more (meta)  
>   _traits power.
>
> --Uwe Salomon's Indigo library is also offering a lot of ideas in this  
> regard .. one thing I would name here is internationalization.
>
> --QT in D.... ?
> D bindings to GUI APIs are already there, but wait ...
>
> /
> I think the point is to give a pure *D2 QT implementation* a go, and  
> let's see WHERE the compiler/language is not good enough.
>
> Like Andrei did it for his std.algorithm.
> /
> Bjoern