On Jul 27, 2011, at 06:46, Andy Canfield wrote: > We're doing a web site. The whole site will be under revision control using > Subversion. > > So there are at least three of these: my working copy, the repository, and > the working web site. > > How do I get a working web site? By export? or by checkout? > > + After we 'svn checkout', If we need to change something on the working web > site, we can 'svn commit' to get the changes into the repository. Using 'svn > export' we would have to make the changes on a working copy, then commit, > then 'svn export' again. > > + After we 'svn checkout', whenever a new version is in the repository we can > 'svn update' and get only the updated files. If we have to use 'svn export' > it will give us ALL of the files. > > - Using 'svn checkout', the working web site will have the subversion control > files in the .svn subdirectory, which might be a security hole. > > Advice?
The .svn directories needn't be a security hole; you can configure your web server to not serve their contents, or even not acknowledge their existence. In my httpd.conf I have: RedirectMatch 404 .*/\.svn(/|$) You should also look into SVN::Notify::Mirror: http://search.cpan.org/perldoc?SVN::Notify::Mirror
