Mark, thanks for the tidbits, they defintely clear up a lot.
Armed with my novice yet growing knowledge, I was thinking about treating
the product web directory as working space and having an svn update script
of some sort croned. The switch options sounds interesting though.
If I used the tags directory as the master for updates, where do developers
committ to and update from? I'm guessing trunks (since tags is copy). My
update process for production could watch either the trunk or the tags
directory, but tags is better (faster more efficient)... is this correct?
2008 at 10:41 AM, Mark Phippard <markphip_at_gmail.com> wrote:
> On Sun, Feb 24, 2008 at 10:33 AM, Jay Smith <jaysmailinglist_at_gmail.com>
> wrote:
> > This is my first post and I am new to subversion.
> >
> > I have a web site..... root dir is /usr/local/www, I created a repo for
> it
> > named www in /usr/local/repo/www.
> >
> > This is where I get a little confused... initially I imported the
> contents
> > of /usr/local/www in to the www repo. Then after reading all the stuff
> about
> > getting the updated files to production (using update, export or cruise
> > control, etc.)... I started wondering how trunks, branches, and tags fit
> in
> > to the picture.
> >
> > This is my current understanding how these directories work, please
> correct
> > me where I am wrong...
> >
> > Trunks, branches, and tags aren't actually directories on my site, they
> are
> > directories in the repo...
>
> Right, and you will not be able to see them by looking in
> /usr/local/repo. The repository is like a database. It is opaque.
> You look at it and work with it using an SVN client.
>
> > After creating the dirs using svn make dir, I would have to import the
> > contents of /usr/local/www into /usr/local/repo/www/trunk...
>
> Right.
>
> > The developers can create branches as necessary, but I would only copy
> > content from the trunk directory to the actual web site...
>
> Branches are for managing the development process. Unless you make
> branches for your releases, you probably would not put them on the
> actual site.
>
> > Tags are like snapshots, so they are the same size as the complete site
> (my
> > /usr/local/www is 5GB), each tag copy will be 5GB... which probably
> makes
> > tags impractical in my case...
>
> Branches/Tags are copies using SVN. A copy takes virtually no space
> at all. It is a pointer. So tag away. The size of your repository
> does not impact the size of the tag.
> .
> > If I use tags, I would copy the content from tags to the actual web
> site,
> > not trunk.
>
> Possibly.
>
> > Using tags make it easier to backout....
>
> Possibly.
>
> Are you planning to export your site or checkout your site? The
> disadvantage to checkout is that the SVN working copy will be present
> which effectively doubles the size. Disk is cheap, that might not be
> a problem. Using checkout gives you huge advantages. When you make
> new changes in trunk or create a new tag for a release, you can use
> svn update or svn switch to quickly and efficiently deploy the change.
> For example, using svn switch to switch your working copy from the
> 1.0 release tag to the 2.0 release tag will intelligently update your
> working copy with the changes between those releases.
>
> --
>
> Thanks
>
> Mark Phippard
> http://markphip.blogspot.com/
>
Received on 2008-02-25 14:05:53 CET