Interesting questions. >From my perspective (if anyone cares :>), the capability to have tag based functions (i.e. being able to return recordsets and such from a function) puts CFMX on par with languages like Visual Basic. The use of CFCs (which are basically functions other applications can call - as I understand them) would be somewhat equivalent to using DLLs to encapsulate common logic.
With that in mind, then future development in CF will likely follow a path similar to Visual Basic, C/C++, Perl, etc. That doesn't necessarily help any CF coders who do not have experience in other (desktop and/or client/server) languages. I can see these people having a bit more of a difficult time - and I'm sure a new Fusebox like methodology will come along, or Fusebox will grow to handle this to some degree. For me, the bottom line is that there is very little change in the development process. 1) understand what the application is supposed to do (i.e. a requirements document), 2) analyse the data required, where to get it, how to manage it, 3) prototype an interface, 4) develop an "architecture" (what are the core functions, where are they placed, how do you call them, etc.), 5) Write the code to make the interface work as expected (which includes all the "behind the scenes" code. This method is the way I've been taught, and have seen it in action professionally. If followed, the projects flow nicely. If ignored, the projects tend to get delayed and have other issues. And this method applies to C/C++ projects, VB projects, Web projects, or any other development project. So, as I said, I see little in the way of change for me. I just have to spend some time to learn how these new CFMX features work, and when to apply them (which would apply to step 4 above). That said, it'll be interesting to see how things develop. I'll be watching the list, and learning what I can. Shawn Grover -----Original Message----- From: Matt Liotta [mailto:[EMAIL PROTECTED]] Sent: Monday, April 29, 2002 8:30 PM To: CF-Talk Subject: CFMX: Version 1 all over again After all the emails about CFMX, I could resist starting another thread. Seriously though, now that CFMX is here, I think it is time to start asking some hard questions about how we develop with CF. While we could debate forever the merits of different application frameworks and methodologies; it is safe to say that most experienced CFers already have decided on one. I know that if I was asked today to build any kind of web application in CF 5, I would know exactly how to attack it. However, I don't think this applies to CFMX. Will old methodologies and application frameworks from CF 5 continue to work on CFMX? Yes, for the most part we some minor changes. Of course that doesn't mean they will behave the same. Do you have techniques that help you build fast performing CF applications? Think those techniques will still work with CFMX? There are all kinds of new things to take into account with CFMX. First, there are some new features like CFCs that fundamentally change the way applications can be put together. Second, some of J2EE's features are exposed to CFMX like JSP:forward and servlet filters. What neat things can you do with those? Finally, existing ideas about performance mostly have to be thrown out and reevaluated. All of the above means it is time to start over as a CF developer and take a hard look at how to build applications with CFMX. Never before has there been such fundamental changes to the platform that affect every aspect of CF development. Don't make the mistake of continuing down the same path you were on with CF 5. Stop and smell the roses and pick the best path. -Matt ______________________________________________________________________ Structure your ColdFusion code with Fusebox. Get the official book at http://www.fusionauthority.com/bkinfo.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
