I've been looking forward to the release of Subversion 1.5 for some time now because it feels like the ability to track merges is the final major roadblock keeping Subversion from being an enterprise-ready alternative to ClearCase. That darn red arrow has represented my last bit of loyalty to the ClearCase solution.
I'm a big pusher in my company of focusing our tool selection recommendations on the actual needs of clients. At times, ClearCase is the right version control solution. In other situations, ClearCase just isn't practical. I like Subversion's ability to mimic UCM's component-based structure in repositories without having to take on UCM's process control and anti-agile methodology. So, I've been a big fan of the Subversion thing and have thought of myself on the leading edge of version control.
Then I read this interesting blog entry by Ben Collins-Sussman and I see how wrong I was. He has some valuable things to say about the 20% of the industry that are uber-geeks fascinated by the most recent shiny thing and the 80% who are just trying to work through their hours. And you know what the 20% are interested in in VC terms these days? It's not Subversion, it's "distributed version control systems." Dang. Guess I'm not in the 20%.
On the other hand, the 80% is still stuck in the old lock and load world of PVCS and ClearCase. So, I'm not over there either.
Actually, Mr. Collins-Sussman is making an excellent point here and it applies to everything I do, not just the version control stuff. We are currently running around playing with the shiny agile playthings (I'm now a certified ScrumMaster!) and yet most of our customers haven't completely bought into the iterative notions of RUP! We're all excited about transparency and progress, but Second Life gets slammed because it adds too many new features.
We can disparage them all we like, but not everybody is all excited about the new shiny stuff.