|
|---|
| Replies are listed 'Best First'. | |
|---|---|
|
Re: User Interfaces
by Masem (Monsignor) on May 01, 2001 at 21:57 UTC | |
The best way to do any sort of problem that might have a graphical interface is to remove as much of the non-GUI-based logic from the code into a module, and then write GUI code that uses that module; this way, if you write a QT app and I'd like to make a GTK version, I don't have to play around with your logic, only do the necessary 'mapping' of GUI elements.
Dr. Michael K. Neylon - mneylon-pm@masemware.com || "You've left the lens cap of your mind on again, Pinky" - The Brain | [reply] |
by Reverend Phil (Pilgrim) on May 02, 2001 at 00:06 UTC | |
*shrug* A lack of a standard is often an indication that a standard needs to be developed, rather than a sign that all movement in the direction of a standard is hopeless. I think it would be rather keen. -=rev=- | [reply] |
by jeroenes (Priest) on May 02, 2001 at 15:35 UTC | |
Such a standard would really be useful IMHO. Jeroen | [reply] |
by petesmiley (Friar) on May 01, 2001 at 22:36 UTC | |
Besides, I thought it wasn't necessarily just SQL that was important, I thought it was an issue of database drivers, which don't all communicate the same. The SQL commands between databases might be the same, but the method for connecting and communicating would be different. I certainly realize that user interfaces can be way more complex than database interfaces in some ways. However, I think the idea is amusing. | [reply] |
|
Re: User Interfaces
by clemburg (Curate) on May 02, 2001 at 16:48 UTC | |
Interestingly, the Xanalys Lispworks Common Lisp Compiler contains such a thing. It is called the CAPI. From the docs: "The CAPI (Common Application Programmer's Interface) is a library for implementing portable window-based application interfaces. It is a conceptually simple, CLOS-based model of interface elements and their interaction. It provides a standard set of these elements and their behaviors, as well as giving you the opportunity to define elements of your own. The CAPI's model of window-based user interfaces is an abstraction of the concepts that are shared between all contemporary window systems, such that you do not need to consider the details of a particular system. These hidden details are taken care of by a back end library written for that system alone. An advantage of making this abstraction is that each of the system-specific libraries can be highly specialized, concentrating on getting things right for that particular window system. Furthermore, because the implementation libraries and the CAPI model are completely separate, libraries can be written for new window systems without affecting either the CAPI model or the applications you have written with it. The CAPI currently runs under X Window System with Motif, and Microsoft Windows."
Christian Lemburg | [reply] |
by petesmiley (Friar) on May 02, 2001 at 19:52 UTC | |
| [reply] |
|
Re: User Interfaces
by jplindstrom (Monsignor) on May 02, 2001 at 10:11 UTC | |
A similar GUI system would be a programming related solution to a user interface design related problem. Bad. Your program might be cross platform. But your users aren't. For example, if they boot Windows they expect a Windows GUI. Actually, the deserve a Windows GUI. Anything else will reflect badly on your program, however unfair that is. The problem with a cross platform system is the details. Everything has to be right in a good GUI, down to the last pixel. Like Masem said, separating logic from presentation is a very good way to go. In some ways it will make your logic more complex (program flow will jump between classes where it previously didn't), in some ways it will make it more clean (things that belong together will be together). It will surely force you to think more about it which is probably a good thing. It is easier if you don't mix UI metaphors (like one Command Line Interface and one GUI). Tk and Win32::GUI generally have more in common than they have with a batch style program. /J | [reply] |