Hello,

I am contemplating shifting from CVS to Subversion (SVN). I tried a test import from my CVS repository to a new SVN repo. Then, I checked out my work into a test work directory. The results I got have me wondering if SVN is really going to work for me.

First, I'm not a code developer, I'm a writer. In my home directory, I have a directory called writing. In that directory there are further divisions: ~/writing/fiction/, ~/writing/ideas/, ~/writing/support/ ...and so on. I usually work in one directory per story and a novel has several directories to cover those tidbits of information I collect that don't go into the novel.

However, in the test SVN checkout directory, I have a totally different picture. Instead of seeing my work, I'm greeted with a chaos of tags and branches. Some of my work is in the trunk directory and some is scattered in tags or branches. Is this right? Is this the way SVN works? I saw FAQ topics fretting the .svn directory, but none that addressed my concerns: why my work is spread across this chaos?

In the meantime, I wonder if there's some other tool out there that would be better suited to helping me track edits and changes in my work. Expecting responses that cover the editing abilities of most word processors: I don't use them. I work in simple, beautiful, plain text (and then bring in the big guns: LaTeX!).

Thanks in advance to any SVN or CVS afficionados or other writers who see this and take the time to reply. This community is my favorite place to come for answers.

Anne.

you can convert to subversion via ::

http://subversion.tigris.org/faq.html#cvs2svn

Importing is problematic at best.

Yep, that's how I got to the point I was discussing in my original post. I've done the conversion (at least a test run) and I've done a checkout. What I've got now in my working directory makes a pile of spaghetti look easy to untangle. This is such a departure from what I expected to see that I wonder if I'm doing it right.

Anne.

I know this is a bit late, but it might help others.

Subversion does split the work into trunk, branches, and tags as you say, but when you do a checkout you specify then which you want. i.e.

svn co url_to_repository/trunk/project

As I'm guessing you know from the documentation the trunk if for the main project. In development terms (I know this doesn't apply as a writer) this is where the latest working version is kept.

The branches are where you can keep variations to the work, and actively work on them. The tags simply provide the equivalent of a symbolic link to a certain revision whether that's in a branch or the trunk. This means instead of having to remember the revision number you can refer to it by a specific name. For example

svn co url_to_repository/tags/sent_to_editor

Once you have the repository set up and running, you don't check it out from the root repository directory so you don't have to deal with branches, trunk, and tags each time.

In your case if you're not going to be experimenting with different versions to your text then simply using trunk all the time, with a tag every now and then to refer to a specific past snapshot might be best? It sounds like you can just ignore branches. Once you've made them you won't ever have to remember that they are there if you don't want to.

Hope this helps,

mehmet