I have a client who I’m trying to convince to implement Subversion (I know, I know). My pitch goes something like this: His test domain is visible to Google and other webcrawlers, despite his best efforts (robot.txt not withstanding). By putting the repository on an svn:// URL, we could avoid having the code crawled until the deployment to the live domain.
My question is: is there a way to deploy to a test domain, then bring it back down after work has been done. I guess what I’m asking after is some means of switching the test domain on and off easily, and through Subversion or some tool in conjunction with Subversion.
This is more of a client-management issue, but y’all have never let me down.
There are many reasons to use SVN, but I don’t think “preventing your code from being crawled” is one of them.
If your development areas and code can be crawled by search engines, you have a serious system administration issue. There are dozens of solutions to this problem… IP whitelists, HTTP Authentication on a per-site basis, per-domain ACLs, etc. HTTP Auth takes 5 seconds to set up… you should really have something going.
The myriad of benefits for SVN are covered everywhere, so I won’t extoll it’s virtues here. Google “why use SVN” to compile a laundry list of good reasons.
As far as test domains go, that’s not really something to be controlling with SVN either. It’s as simple as a quick web server configuration change to determine which sites are live and which aren’t. However, if you had controlled access to your test domain, you wouldn’t even need to do this.