At 09:47 29/11/00 -0600, you wrote: >> Yep - if it was setup well. But it would take a *lot* of work to make it >> worthwhile so ... ;) >> > >What part of it would take a lot of work? Don't you just need to >configure a list of URLs, something you have to do anyway to keep the >repository up to date? > >What am I missing?
To work and keep it easy you are going to have to do a number of things. * Maintain clientside of jar versions/locations/dependencies etc * Have some way for regular checking to see when new versions are updated such that anytime a jar is added to project something will communicate to CVS clients that an update is needed. (And no relying on humans is not acceptable). * Have some way of all committers being able to change dependencies etc. (Possibly by incrementing micro-version) and still maintain a version history * Work out a method for using behind firewalls * way of incorporating local libraries into database (and thus not forcing a redownload every build) etc. There is a lot of minor issues about that can not be adequetly solved with the current infrastructure of java apps. While there is people who actually use the Java recommended versioning manifest entries the vast majority don't know about them or don't care about them. Have a look at some of the lists for RPM/APT-GET/InStallshield type apps and you will see that while it sounds like a real easy job - it isn't really ;) Cheers, Pete *------------------------------------------------------* | "Computers are useless. They can only give you | | answers." - Pablo Picasso | *------------------------------------------------------*
