> ... If there were a single XUL standard then IDE developers would be motivated to create forms designers for it?
A standard doesn't imply that something will be used. It also has to work with the existing tool base. > ... What does that mean? Why would I want MFC in a cross-platform GUI? Why would you WANT a cross-platform GUI? My experiences are probably different, but the cross-platform applications that I've seen, that try to unify *anything* except low level code, fail miserable. The GUI's are non-standard, you can't use the usual tools during development, and the overall quality of the application suffers. Sad as it is, if I want to develop a cross platform application, I'll develop separate apps for each platform so that users get a familiar-for-their-platform look & feel and the performance of the app is tailored to the platform. > ... If you want that, you can have it: http://www.mozilla.org/projects/xpcom/ Hmmm. I meant interfacing to COM objects written around the "Microsoft" implementation. Looking at this website, I don't really see that the support for that is there. It seems to be its own custom COM implementation, but I may have missed something in my brief perusal of the website. > ... It does not make sense to conflate "security" with "binary". I won't bother addressing this point because it is nonsensical. OK, maybe I didn't state it right. I've been told by many people that they would not ship commercial applications with the markup in human-readable, easily changeable form. They want the user interface to their application "secured", if I may use that term. > ... And runtime conversion to native code is a lot slower than compile time conversion but that's the direction Microsoft is going with C#. >From my understanding, Microsoft's implementation is 1) a serialization of the form layout, and 2) would normally be used to generate native code at compile time, not run time. The fact that it can do run time code generation does not mean that you would want to do that, especially if performance is an issue. It's there if you need dynamic GUI generation. Even then, once the markup has been created, generating native code and compiling that, sure, is slower the first time. But as long as the markup doesn't change, or changes every so often, then you can use the runtime generated version, at least until the markup changes again. I think there's an advantage in that approach. > ... ActiveState has built a proprietary application on Mozilla's XUL. OK. Still, I've gotten a lot responses regarding concerns over an open source effort. Not just licensing, but support, changes, etc. > ... Mozilla is all C++. OK. > ... XBL. So? Why would anyone want to write in XBL (or any other markup) directly? If you want acceptance of a standard, if you want to set the trend, then you can't just waive your flag and say "I have a standard". You need tools for developer acceptance. A migration path for existing code base. Products that work not just equally as well as existing ones, but better. And that's really my point--the XUL community needs to get together and get behind a suite of tools that demonstrate to the developer that there's a viable alternative to the "Microsoft Way". Is Mozilla that alternative? I personally don't think so. > ... Do you think that .NET is the only way to do dynamic loading? No, of course not. I've implemented several dynamic loading frameworks in C++, and there's numerous other plug-in architectures out there. > ... And what is it about .NET reflection that is so much more interesting than either Java or Active-X reflection? Good question. The answer to that is difficult to nail down. Personally, I don't see Java as a widely accepted language that meets developer needs across several key environments. And I'm not sure Active-X is embraced that well either. One of the "problems", trying to get back on track with the XUL issue, is, for me at least, in the assumption that cross platform development capability is a good thing. Frankly, that's not a given for me. There are niches, of course, like browsers (which in themselves don't really hold together as a standard, given the number of "if browser=this else if browser=that" cases I see in JavaScript). However, in general, if XUL is being sold on the idea that it allows cross-platform development, then I think that's the wrong, erm... platform! Marc ------------------------------------------------------- This SF.Net email is sponsored by: IBM Linux Tutorials Free Linux tutorial presented by Daniel Robbins, President and CEO of GenToo technologies. Learn everything from fundamentals to system administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click _______________________________________________ xul-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/xul-talk
