[RESOLVED] Version Controll

jmartinez

I need a clean way to keep all versions of a group of PHP files archived and versions updated. Currently all I do is backup the most recent version and begin to recode/add/remove files, but there must be a more accurate way of controling this information... right? I have looked at CVS stuff and am completely lost.

So my question to you is:

"how do you keep track of your php file versions for your projects?"

I work with Zend studio, on a Windows Ultimate OS, and am running WOS Portable with PHP5_SE (Version 5.2.3) and MySQL5_SE (Version 5.0.41).

Any suggestions would be greatly appriciated!

Thanks in advance!

Josh

laserlight

I have looked at CVS stuff and am completely lost.

Read some documentation and tutorials. For example, there is A CVS Book. Chances are TortoiseCVS, which integrates CVS into Windows Explorer, will work for you. I do not really know as I use Subversion's TortoiseSVN when I have to use Subversion (and then there's the fact that I am on Windows XP).

how do you keep track of your php file versions for your projects?

I use the Bazaar version control system from the command line as I found that easy enough, though there is an experimental TortoiseBZR along the lines of TortoiseCVS. Of course, you still cannot avoid reading the documentation.

MarkR

I use Subversion. You must however, understand how to use it correctly if you're to achieve anything:

Each developer MUST have their own completely independent development environment (even if they are all on the same dev server) including database.
NO code must reach your production server without going through version control - developers must not be allowed to simply copy files from their dev area into production.

You will need to consult some documentation as to a typical lifecycle. Normally each developer will work in their own area, committing their changes when they're happy and merging one anothers' changes as they feel like it. At some point, when a release is imminent, all developers will stop committing any new changes while system testing is done (i.e. a freeze) - developers might continue to commit to their own branches though.

Once testing is complete (ideally have another area or server for this), you can then take the same package of files that you edited and deploy it into production by some organised mechanism - ideally do so in a way that deleted files are also removed.

Configuration and data files need to be excluded so they're not accidentally overwritten during this process.

Database schema versioning is quite tricky- you will probably find some examples of how it's normally done - all development and production databases need to have a schema that's in sync with the code base they're running at that particular moment (of course they won't all be the same).

Mark

jmartinez

Thanks guys, I actually started to use activeCollab 0.7.1. I am the only developer for the company I work for, and will be for quite some time so this is starting to work pretty good for me. I can later add more "users" either other developers or clients if I so choose. It is a simple format for me to understand. All I do is upload and comment on the latest revisions to a specific file for that days work on it. Simply put, its just a way for me to keep track of what the heck I did yesterday. Thanks for your input!