What Programming Languages Do You Struggle To Hire For?

This is exactly what we do, and it's worked really well. Of course, we're doing Agile, which I think helps this process.

But if I have a good candidate that has attention to detail, knowledge of basic concepts, etc. but doesn't know our language of choice, I'm confident we can teach him our language. Especially if he has already learned similar languages. Even if he hasn't, we've hired people with almost no experience as programmers but that have the ability to absorb knowledge and have a passion for wanting to be a programmer.

Then, we've also had applicants that could "talk the talk" and had experience, but ended up being terrible programmers because they lacked attention to detail, or didn't want to be bothered with writing unit tests, or whatever.

Obviously, I'd much rather have someone that I could train with good habits than someone who has experience with bad habits.

For loops.

True, but I've seen lots of crummy in-house "custom libraries, obscure frameworks, legacy code, [and] tooling" that would have been better if they had been written by veterans of the target language. Or better yet, the veteran would know how to apply an off-the-shelf technology to solve the same problem.

Exactly. I can get a guy up to speed with C# in a matter of hours. The problem comes up when we've got to deal with issues in IIS, Entity Framework, ASP.Net MVC, etc. etc.

Most commodity programming languages are simple. Either they feel kinda like Javascript, or they feel kinda like Java. Obviously there are a few shops out there doing elaborate stuff with Clojure and they have a higher bar, but probably 95% of the work out there is being done with "typical statically-typed VM-based language" or "typical dynamically-typed language".

I'd like to come at it from the opposite direction, and help engineers build out profiles of what they are looking for. Not another resume site, but a forum for making anonymous declarations of what the the perfect job(s) would look like. And something similar for potential employers. I probably wouldn't even focus on making connections, but instead just provide scores and feedback on availability, trends, regional factors, indirect costs, etc.

It seems like that sort of data would be worth something to new and established businesses, who both have to sort through an increasingly crowded forest of tools and technology.