Subversion for web projects?

[Superwormy]

What I'd like to do is use Subversion or CVS to help keep archived copies of all of our code for our website.

What I'm confused about is it seems like all code in the subversion repository is in some special form in some special directory structure... how do I get those files *OUT* of subversion to be able to test them / compile them / have apache edit the PHP scripts?

Right now all of the code sits on a development server in the /usr/local/www/ directory, can I do something like make /usr/local/www/ a symlink to the latest version of all of the files in the subversion repository or....? How does this work?

Thanks so much!

 

[n0cmonkey]

If svn is anything like cvs you have to check out the files.

 

[Superwormy]

So does that mean to get a working copy to test with I have to check out all of the files each time I make a change?

So like I'd check out the entire project each morning to a seperate working directory ( /usr/local/www/ ) test, test, test, make changes, test more, etc. and then check back in the entire proejct that evening when I leave work or...?

 

[Superwormy]

Anyone...?

 

[n0cmonkey]

You don't have to bump so often. You should be able to check out individual files, or even specific revisions of files.

EDIT: No need to bump so quickly. 4 hours or wait till it's off the front page.

 

[Superwormy]

Right I understand that I can check out individual files, but to be able to test I need to check out probably half of the files of the project... so my question is do I need to check out the entire project ( essentially make a copy of the entire project ) everytime I want to test something? Or is there a way to just directly get at the latest version of all of the files in the repository and have Apache serve those files...?

Sorry about the bump.

 

[Drakkon]

yes you would have to make your own copy of the files each time you want to test
I use tortoiseCVS ( http://tortoisecvs.org/ ) to create folders that link to the CVS within my apache folder so i dont have to work from some other wierd folder

 

[Armitage]

The way I do it - I create checkout two working directories of the entire project. One is for development & testing, the other is the "live" version that is actually out on the web/used by others/etc.

Do al your work and testing in the development directory. When you're happy with it, check it in (commit). Then go over to your live directory and do an update - this will pull the changes you made into the production version.

 

[Superwormy]

Thanks Armitage, thats what I needed to know, thats very similar to what we do now, only minus the CVS/Subversion bit.

Thanks nOc and Drak too, though server is FreeBSD and development is in Eclipse, so Tortoise isn't an option really.