[erlang-questions] Time for OTP to be Renamed?

Vance Shipley <>
Sun Feb 16 10:13:44 CET 2014


On Thu, Feb 13, 2014 at 12:22:54PM -0600, kraythe . wrote:
}  Not so much the issue of OTP as a name but the general impression of
}  "adoption? bah humbug ... if they don't like it they all suck and we
}  don't care." Sounds like a disaster of a business decision to me.

Erlang has always been a very practical thing.  It is a functional
language, but not purely so.  It's developers have always chosen the
path which made sense for it's commercial use.  It's not a language
made to teach computer science courses, it's made to be as useful as
possible to build real life commercial applications.  It's rooted in
industry.  You're questioning the business decisions of a company which
has been at this for over two decades.

}  Im not so sure its going to be easy to staff or finance a project on a
}  language that has 1) tools that need work, 2) a limited trained staff 3) a
}  community that cares little about language adoption. It could be that
}  Erlang becomes another Lisp for me. A language I think rocks but is
}  entirely impractical in the business world.

My greatest wish is that any potential competitor of mine believes all
of the FUD above.

The truth is (my competitors shouldn't trust me, I'm probably lying) that
Erlang/OTP is a very practical platform for many business applications.
It provides a rapid development environment for reliable and scalable
applications.  It's battle hardened in the demanding communications
service provider industry.  If you're building cloud services or other
massively concurrent applications you will get a lot of stuff for free.

There is a barrier to entry here, but it's mostly a phycological one.
Yes, you need to learn about functional programming and OTP however
at the end of the day you're still much, much further ahead than if you 
start writing the code you are productive at today and reinvent (poorly)
much of the functionality of OTP.  A few months in and you have learned
how to implement new business problems quickly and easily at scale.

For the most part I don't hire people for what they already know (beyond
computer literacy).  I hire them because they've demonstrated an ability
to learn.  I'd rather take on someone who has learned a bit of many
languages than the master of one.  A good hire is someone who considers
themselves a software engineer and not a Java(/whatever) coder.  It's
true that many employees will be concerned about their resumes and a
few years of Erlang experience may not seem that marketable to them. In
reality though these are the folks who have less value us.

... and speaking of Lisp, this is a great read:

	http://paulgraham.com/avg.html

-- 
	-Vance



More information about the erlang-questions mailing list