cl-blog fun
I've synced up what I'm using on this blog and what's in SVN, and tagged the result as stable-maybe-1.0. Of course SVN's concept of a tag is somewhat different than CVS - in SVN it's common to keep your "HEAD" in a directory called "trunk", and tags are just copies of that part of the repository in some other directory. Accordingly I've made the following modifications to rearrange the repository:
- The latest and likely now-broken greatest is at svn://unmutual.info/cl-blog/trunk/cl-blog/
- The stable-ish version is at svn://unmutual.info/cl-blog/tags/stable-maybe-1.0/cl-blog/
What's in that now-broken trunk version? Well, I've moved the serialization off into its own macro called define-simple-serialized-class, which manages a list store of items that are serialized to a directory with explicit serialization commands. References to other serialized objects work OK too. Also I've been experimenting with using templates for as much as possible, and would like to make cl-blog more templated if I can.
Of course, I just took what's on my hard drive and committed it, so I'm not actually sure if what I just committed actually works - more testing will come. At some point I'd like to split up my blog into one blog for technical stuff and another blog where I can wax philosophic or political from time to time if I want to. The more technicalish blog would also serve as my web page, then, instead of the old cobweb I have laying around on an account at a school I'm no longer at.
It may be that all the cool kids are using darcs, but then I've never been much of a cool kid. For the most part I'm the only one doing serious hacking on my projects, and having more than one person doing so without sharing a repository just seems like a recipe for confusion to me. While darcs is in my experience the easiest of the distributed SCMs to pick up, it's still not quite what I'm used to. I am quite happy with SVN, especially when I discover little things like this:
The svn switch command transforms an existing working copy into a different branch.
This meant that after I reorganized the cl-blog repository along the lines that I described above, I took the working copy of cl-blog that was running this site and switched it to the new URL. There were no changes fetched by svn - it realized that this working copy was what I had committed to the repository from that directory, then moved into trunk/cl-blog, then copied into the tags/stable-maybe-1.0/cl-blog directory. I'm not even sure how that works since it's really just saying "replace this directory with this other directory in the repository", but as long as it does I'll be happy.
Addendum: Wonder of wonders, what's in the trunk works. Now to keep working on it...
One addition to the above: one reason I like working with SVN so much is the SVN Book, which is a great reference and tutorial book and makes it really easy to adapt from using CVS.