Ok. Maybe it's useful to have a devil's advocate position here. There's often a form of argument that says there's really not much of a difference between languages, tools, etc., that it's simply a matter of taste any anyone could be just as productive, successful, happy, etc., regardless of the actual tool choices. I'm going to say that this is usually a way of avoiding thinking about the problem and dismissing the opportunity costs by not examining them.
Version control systems do interact with languages. If you we're working with an IDE oriented language like Java you might be limited to whatever VCS works with your IDE. Since you're working with Ruby, which isn't typically developed using IDE refactoring tools, you have more freedom. Many people choose their VCS for these kinds of legacy reasons. It doesn't sound like you have to.
Choosing Subversion for Ruby, while it certainly won't kill you, I think is a missed opportunity to explore more modern development environments and ways of working with others. For an idea of this, here's a talk that's well worth listening to: http://www.youtube.com/watch?v=4XpnKHJAok8\. Torvalds uses strong language when he discusses Subversion, as is his style, but he's bright guy, he's been extremely successful in managing important software development projects, he's spent a good amount of time thinking about the problem, and he's far from alone in his thinking.
Whether you choose something like Git, Mercurial, Bazaar, or even Darcs, I think you'll end up learning some useful concepts of growing importance that you can't learn with CVS or Subversion and it won't cost you anything.
···
On 1 oct. 08, at 11:57, thiagobrandam wrote:
Thank you guys. I didn't know my doubt would generate such rich
discussion. This semester I'm having OOP classes
and as a final exam, we will have to build a software using OO
concepts. As this is pretty much the first time my group
will have to deal with revision management. We decided to program in
Ruby (solely, no Rails at all) and we were wondering if there was
such thing
as a good revision control tool that best fits our needs (turns out it
doesn't according to the discussion...). We talked to our teacher
and he decided that the Subversion repository of Google Code would be
the standard for all groups. Thanks once again for all the help. I
found
Robert Klemme's and David Masover's opinions particularly interesting.
Help us think about hypeXreasoning issues and focus on what really
matters.
I generally use git for all my projects and use git-svn when I need to
pull something from or publish something to our centralized corporate
subversion server.
···
On Wed, Oct 1, 2008 at 11:57 AM, thiagobrandam <thiagobrandam@gmail.com> wrote:
We talked to our teacher
and he decided that the Subversion repository of Google Code would be
the standard for all groups.
I'm drinking deep from the Git Koolaid, but it always surprises me to see how often this video is recommended. It's just terrible. Linus doesn't really say anything of value in here. Well, to be fair, his "circle of trust" tangent is very interesting, but has little to do with Git.
Anyway, my opinion is that Git is great and you need to check it out, but this video is entirely skippable.
I have to disagree. It isn't concise, but there's a lot of insight and history here. Also, unlike many of the more theoretical discussions, this talk ties into one of the world's most successful software development projects and the importance of version control with respect to its development. It may be that you have to look below the surface a bit. Notice for example, that in the early days Torvalds used tarballs and diffs instead of the dominant open source tool CVS. A shockingly choice, but less so if you're familiar with the design of CVS and the requirements of the project.
Anyway, the point of the talk isn't git. I didn't propose it as a pitch for git specifically. If you read a bit between the lines you see that the talk is about how software development and version control interact. But if you have other discussions you think are really good then please post them.
I'm drinking deep from the Git Koolaid, but it always surprises me to see how often this video is recommended. It's just terrible. Linus doesn't really say anything of value in here. Well, to be fair, his "circle of trust" tangent is very interesting, but has little to do with Git.
Anyway, my opinion is that Git is great and you need to check it out, but this video is entirely skippable.
I'm drinking deep from the Git Koolaid, but it always surprises me to see
how often this video is recommended. It's just terrible. Linus doesn't
really say anything of value in here. Well, to be fair, his "circle of
trust" tangent is very interesting, but has little to do with Git