Mathieu Bouchard <matju / sympatico.ca> writes: > On Thu, 24 May 2001, Mario Lang wrote: > > > What I am thinking about is a Interactive Ruby IDE. > > > But the interface indepence bothers me a bit. I cant find out > > how to structure a program from the grounds up well, so > > that different interfaces can be written for it later on without > > modification of the core. > > Let me define both parts, Model and View. > [.MVC.] > means the Model knows the View, but only in the way that an Observable > knows its observers. When the View decides to update what it presents to > the User, it queries the Model about its properties and presents that > information. As long as their is no coupling between the model and/or the view, I like it. > Now I didn't say what is between the View and the User. We could call it a > Display and that Display needs not to be visual; it could be X11, NCurses, > a command-line interface, a voice-based system, anything. For a given kind > of Display you need a set of Views that make a bridge between the Display > and a fair amount of Models. Yes. Someone got my point! Because I am not talking about independence between Qt and GTK, I am talking about real interface independence. No matter what interface. Yup, their are still people thinking like that. Thats a nice thing to begin a day with :). > > Perhaps some project like this is already on the way (the ruby IDE) > > I don't remember who planned to do one besides me but I haven't heard of > a serious initiative. > > > If this apllies, I would be happy to join forces. And I would hope to > > convince the author of the importance of interface independence. > > You won't need to convince me. Good to hear this. > > I have also found the textbuf extension, which I really liked. > > My hope is that MetaRuby makes it easier to experiment in efficient > representations of strings. I'd like that loading a one meg text file in a > String results in a thousand actual Strings all linked together through > StringMixin. Hmm, I read about MetaRuby some times in this newgroups. But everytime someone used the word, he expected every reader to know what it means :-). Does anyone have a link which explains what metaruby is supposed to be? > This is doable now but I'm not so hot in efficient data structures > so someone else would have to pick up MetaRuby and program such a > String class. > > > Example, currently someone is writing the table.el extension for Emacs, > > which allows editing of tabular data. He uses many tricks to > > If you want a "buffer" really based on a concept other than a String, then > probably that making it as general as a Model (see above) is a good idea. That is what I thought about. It should be abstracted from the beginning on, and one actual inherited buffer type will be a string. But I would like to allow it in the future to derive other different buffer classes so that people could e.g. write small applications just like Emacs users do. > I'm very interested in stuff like the Outline mode of Emacs. Yes, Emacs has so many nice concepts. ANd I am sure if they are applied and reimplemented in a pure object oriented sense ... > > BTW: Does Rumacs sound good to you? :) > The name I have found for it is "Hybris". Sounds nice too. -- CYa, Mario <mlang / delysid.org> Homepage(s): http://delysid.org | http://piss.at/ "How should I know if it works? That's what beta testers are for. I only coded it." (Attributed to Linus Torvalds, somewhere in a posting)