Revision 1 as of 2005-03-22 05:48:35
missing edit-log entry for this revision
converted to 1.6 markup
|No differences found!|
Which is correct:
a) Grab commons jars from most recent struts nightly b) Build commons jars from commons CVS c) Grab commons jars from most recent commons nightlies c) The "right way" that I have not yet stumbled on....
What exactly does the struts nightly build do? Dates inside commons jars seem to vary.
Build.properties.sample seems to expect b) or c). Does anyone have bash or ant scripts that automate the commons prep?
I think c2 would be b, since the idea is that Struts developers are often Commons developers too and would have the Commons CVS checked out and built. So the "ideal Struts developer" would have a current build of the Commons stuff and would run the Struts build against that.
If you don't need the Commons code checked-out, a good approach is to use a "repository" system.
- Download whatever dependent JARs you need to a common directory.
- Edit the build properties file to use that directory.
The current Commons JARs can be downloaded directly from < http://gump.covalent.net/jars/ > so you don't need to grab the whole gorilla if you only want the banana. =:0)
(The Gump JARs ink is down right now. I'll check on that.)
There are project management systems (like Maven) that can snag the JARs for you as part of the build process. When we get past 1.1., I think we will be looking at some likely candidates for this (Maven, Forrest, Centipede).
I used the repository approach for some time, though now I'm moving back to b. It's quite helpful to have all the source on hand if you are stepping through the code in a debugger.