"the business will be able to find someone who is willing to learn in their spare time. Besides, the great thing about being a programmer is not that you have to learn new things but that you get to."
Huh. I love to learn new things, but I hate learning boring technologies that re-solve solved problems. Especially when the old, boring technologies are serving me well.
If I could reclaim the hours I spent learning git (which is only a marginal net improvement over svn), I have a long backlog of CS papers about everything from image classification to distributed systems design. Those problems are interesting. Learning another version-control system is not interesting (and the same goes for: editors, make systems, debuggers, and to some extent, programming languages. I don't want to spend hundreds of hours learning another syntactic variant of Algol or Lisp unless I need it for something.) But now that I'm proficient with git, I can guarantee that some obnoxious fanboy is going to come along and force me to switch to some whizzy new version control system that promises to solve all my problems and give me a pony. It's the way of the world, but I'm still waiting for my pony.
When I was younger, I'd get distracted by any flash in the pan. Now I'm more conservative about what I'll pick up. Maybe that kind of conservatism is part of what you're paying for when you hire an experienced employee. After all, someone who won't waste hundreds of developer hours switching your whole company to the newest flavor of version control (without sufficient justification for the cost) is a benefit to the bottom line.
I don't know if you can draw that drastic of a conclusion from his post without some other knowledge of his particular situation. For some people Git offers very little advantage over SVN, a small team with a single desktop or embedded produce is not going to see the same advantages of a DCVS system that a massive web team does that is seeing daily releases. As such a developer with their use case may feel that the effort to learn it for their situation was not worth the effort.
"For some people Git offers very little advantage over SVN, a small team with a single desktop or embedded produce is not going to see the same advantages of a DCVS system"
I disagree, actually. Small, informal teams can use pretty much whatever they want with little penalty, and may actually benefit from having distributed repositories. Larger teams rarely need distributed version control (there's almost always a canonical central repository), and are also the ones most penalized by the complexity of git.
Git is a complex, confusing beast with a high potential for mistakes. There are some nice things about it, but (from personal experience) it also tends to create as many new problems as it solves. These problems are amplified on large teams.
Huh. I love to learn new things, but I hate learning boring technologies that re-solve solved problems. Especially when the old, boring technologies are serving me well.
If I could reclaim the hours I spent learning git (which is only a marginal net improvement over svn), I have a long backlog of CS papers about everything from image classification to distributed systems design. Those problems are interesting. Learning another version-control system is not interesting (and the same goes for: editors, make systems, debuggers, and to some extent, programming languages. I don't want to spend hundreds of hours learning another syntactic variant of Algol or Lisp unless I need it for something.) But now that I'm proficient with git, I can guarantee that some obnoxious fanboy is going to come along and force me to switch to some whizzy new version control system that promises to solve all my problems and give me a pony. It's the way of the world, but I'm still waiting for my pony.
When I was younger, I'd get distracted by any flash in the pan. Now I'm more conservative about what I'll pick up. Maybe that kind of conservatism is part of what you're paying for when you hire an experienced employee. After all, someone who won't waste hundreds of developer hours switching your whole company to the newest flavor of version control (without sufficient justification for the cost) is a benefit to the bottom line.