Capistrano’s convention is to use an internet accesible server for
your Subversion repostiory. To be honest, having SVN running locally
on your machine is not the best idea from the standpoint that if your
computer dies/blows-up/goes-away, not only have you lost your
development environment, but also your repository.
If you are deploying to the same server that has your SVN repository
(sometimes the case with dedicated hosting), you’d be fine to use
file:///path/to/svn/project – you have to think of capistrano as just
a tool that will go out to your server and do things for you.
So imagine capistrano as an “actor” and when you tell him to deploy,
he will go out to each of your servers (usually one, when you’re
deploying app, web, db) and get your latest code (from wherever you
tell it). This is the repository variable – where the latest version
of your code is.
Capistrano is easy to understand once you’ve done deployment once
manually: sshed into server, checked out code, setup web server,
setup db, then started your app servers and web server…etc. I say
this only because capistrano does all that – it helps to do it
yourself the hard way first.
If you are still wanting to keep your svn on your computer, you’d have
to have a way for when capistrano’s actor sshes into your server where
you’ll be deploying your code, and get back to your computer (if you
are at home, this might mean dynamic dns, and some firewall rules to a
Hope that helps,
Charles Brian Q.
highgroove studios: www.highgroove.com
slingshot hosting: www.slingshothosting.com