Before we check the initial code drop into Subversion, we should
probably talk a little bit about how the repository should be laid
out.
A few questions come to mind:
1) Should the website live with the code, or separate from it?
2) Should we just start off with what we have sitting in trunk/ or
should we assume we will have other language implementations
eventually, and start out with java/trunk instead?
Personally, I'm leaning towards something like this:
abdera/
java/
trunk/
*current code goes here*
tags/
branches
site/
trunk/
*initial website goes here*
tags/
branches/
This leaves room for other languages later without having to rearrange
things, and makes it easy to make tags/branches of one part of the
tree without having to worry about the rest (say we want to release
the java stuff, but not the fledgling C implementation).
It does mean that people will need a separate checkout for the
website, but I don't think that's an overly huge problem...
Comments?
-garrett
