I have responsed to some this before, but here goes. 1. Use Glade to create and maintain the XML GUI interface descriptions. 2. Use libglade to parse and create the interface at run-time and you take care of writing/maintaining your event handlers and other such stuff. This process keeps the GTK Gui stuff cleanly interfaced to and separated from the rest of the code (the really difficult parts of a project)which you or someone else MUST write and maintain. This formula is almost bullet proof when jumping from Windows to Linux or back. -----Original Message----- From: gtk-list-bounces@xxxxxxxxx [mailto:gtk-list-bounces@xxxxxxxxx] On Behalf Of Tristan Van Berkom Sent: Saturday, April 14, 2007 9:58 AM To: Richard Boaz Cc: GTK dev list Subject: Re: How do you develop you GTK code? On Sat, 2007-04-14 at 15:00 +0200, Richard Boaz wrote: > Hi, > > > Like the others posting so far, I do not use any IDE to program my gtk > + app. I had a quick look once upon a time, but quickly determined > that if you want or require to change the code that is automatically > generated, you might as well just write your own instead of trying to > get in the deep-end of someone else's ideas about how this should be > done. Right, first of all I dont think its correct to suppose that generated code is anyone's idea of "how it should be done", generated code is a mechanical spew of "code that works" and should NEVER be used as an example or reference implementation. My opinions about glade are biased because I maintain glade, but my opinion anyway: glade is not an IDE, its a user interface designer - adding application data and modeling of that data is not a desirable thing, you want to keep your user interface in one corner and your data in another - not mix them up in one big pot of spaghetti code. If you have concerns about dynamic portions of your interface, I agree they should mostly be done by hand, but wait - glade can still help you here, say you have a hbox with a label and an entry as was previously described, you can always put one instance of that in your glade file and loop like so: for (i to ... dataentries) { xml = glade_xml_new (..."dataentryitem"...); widget = glade_xml_get_widget (xml, "dataentryitem"); /* ... pack into your program somewhere by hand ... */ g_object_unref (xml); } This might or might not make sence depending on the application, but it does have the advantage of keeping your UI dynamic, if you need to add an item/label/separator or whatever to that subhierarchy than you dont need to edit or compile code to update your interface. I also believe that following a template design might be good for a dumb little CLI wrapper in a GUI - but if you are really developing an industry grade application then please use your brain and understand the code - using libglade for your interface is a good thing because it defers all the repetitive rigid GUI building code to a library that takes care of it for you - it also allows you to use YOUR OWN design that is appropriate for your own application, either by loading dialogs on the fly or at initialization time - keeping pointers around for the widgets that you need referenced by name, or by using lookup_widget() types of paradigms - point being - you are the author of your application and the sources are in your hands, you should be able to organize your code in a way that suits the application at hand. For a good reference implementation - I've repeatedly been pointing people in the direction of the DevHelp sources, since IMO they use a sound OO design and a very practical use of libglade (its also very simple code to understand). Just my 2 cents, cheers, -Tristan _______________________________________________ gtk-list mailing list gtk-list@xxxxxxxxx http://mail.gnome.org/mailman/listinfo/gtk-list _______________________________________________ gtk-list mailing list gtk-list@xxxxxxxxx http://mail.gnome.org/mailman/listinfo/gtk-list
