Hi all,
I'm working on this project which has several dependencies that aren't
in central. A few are the usual Sun suspects, one is jamon 2.7, and
the rest are Internet2 jars which we *could* deploy to central. I
haven't yet brought this up with the I2 people, but I will. In the
meantime, I'm distributing the jars and giving a script to install
them into the user's local repository.
But the big problem is that we're database neutral with respect to our
data store. People in higher ed variously use Oracle, MySQL, MS SQL
Server, Postgres, and no doubt others.
I want to make a binary assembly that anyone can download, drop their
jdbc jar in, and go. This isn't hard.
What I'm dithering about at this point is how to handle the source
distribution and/or people who check the source out from CVS. (Yes,
they're using CVS.)
I can make an assembly that builds a lib directory with the
dependencies and anything in a "user-database" directory.
I can make profiles that will handle the common cases, except for
Oracle and MS SQL Server, but those profiles, if selected, override my
default profile which has an assembly that builds a dependency
directory for running the software in place.
My expectation is that most sites will download the binary
distribution, but some will download the source because they want to
mod it. I don't want to make their lives difficult.
So should I just always include the contents of a directory in the
dependencies in my assemblies, or should I have them check their jdbc
jar into their local repository and edit the pom to have a profile in
case I don't have the version that they want?
Or is there some other way?
I want to do this the Maven way, but I also want it to be reasonably
usable by non-Maven people.
-K
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
