On Sun, Feb 10, 2002 at 09:53:16PM -0700, Chris Fedde wrote: > On Fri, 8 Feb 2002 23:05:57 -0500 Rocco Caputo wrote: > +------------------ > | On Sat, Feb 09, 2002 at 01:52:52AM +0100, [EMAIL PROTECTED] wrote: > | > similar as they would have been chosen to describe objects. So I think > | > it should be possible to transform other notations to UML or whatever is > | > selected. There should be consensus on the features, though. > | > | That would be nifty. Matt's original request was for a feature > | consensus so he'd have some guidance. > +------------------ > > [...] > > +------------------ > | Brian Ingerson has asked me to consider using YAML (www.yaml.org) with > | POE. Maybe that can be the common intermediate language that Your > | Favorite ML is translated into before being passed to a common code > | generator. > +------------------ > > I'm confused. You and Torvald seem to be talking "apples and autos" > here. IIRC the UML is a group of notations used to describe object > relationships, timelines, and behaviors of interest to CASE tool > vendors. If I understand what I've read about YAML, it is a data > serialization language similar to WDDX or Data::Dumper. Is there > some relation between these two ideas that is too advanced for my > pedestrian intellect?
My idea was that Your Favorite Modeling Language would be distilled into some form of common markup language. A back end program can turn that common markup language into code, or maybe just load and run it directly. In this sense, UML (or your favorite CASE tool's language) acts as source code. YAML would be object code, and the generated perl programs would be executables. It's the same concept as compiling .c to .o, and then linking .o to .a. Maybe I don't understand. I've always seen UML as a graphical thing, like flowcharts, and in need of some common textual representation that can be parsed by a single common tool. -- Rocco Caputo / [EMAIL PROTECTED] / poe.perl.org / poe.sf.net
