|Think about Loose Coupling|
Handling version numbersby djw (Vicar)
|on Jul 20, 2001 at 23:45 UTC||Need Help??|
I have started my first "large" project that I am managing on SourceForge.net.
Other than Perl, I have no education or experience in programming so I wanted to ask how people use versioning around here.
Being a linux person, I see most programs come in the form of:
So of course I wanted to use the same format for my program. I started my initial version at 0.3.0 (sounded good - no real reason) and went up from there. I increased the version # depending on how much work/bug fixes went into a release. I am trying to work in the style of "release early and release often" and so some of my updates were small bug-fixes and got an increment of 0.0.1. Some of the other releases that incorporated something new, or made my program work better, got a larger increase like a 0.0.5 increase. Sometimes I was just really happy with what I had changed/added that I bumped up the version number a little farther.
So basically that is how I'm handling it. Really I only have one file that I update and its not all that complex. I'm not in a business atmosphere and I don't have schedules etc for anything.
Now I have heard some things people use to identify their program and that gives their version structure throughout the product life. For example - if the last number is an odd number, that means its a beta. Or if the last number is a 0, then it is a major release.
What are some methodologies used when building your versioning system? Do you use the same system for everything you do? Is there a guideline you like to follow if you can?