Language Bindings for Erlang Again (Opinion)

[Yariv Sadan]

> This is what I consider a real Erlang problem: it is
> fairly obscure. I'd furthermore expect things are
> worse in the US. (Some consider that a competitive
> advantage, of course. Not I.)
>

I agree it's a problem, mostly because popularity often brings many
useful 3rd party libraries and tools. However, I think this can be
fixed in a relatively short amount of time (2 years, roughly the time
it took Ruby on Rails to explode in popularity -- check out
http://www.google.com/trends?q=ruby+on+rails%2C+erlang&ctab=0&geo=all&date=all).
I seriously think Erlang could become quite popular if

- it had a solid web development framework that made common tasks easy
for PHP/Ruby/JSP developers who are "making the switch." (The bar
keeps getting higher, btw -- check out
http://code.google.com/webtoolkit/.)

- it had a friendlier website with community features such as
comments/forums, a wiki and an official blog, which I think are the
best modern-day PR tools for geeky stuff :) http://rubyonrails.org is
a nice example.

- Mnesia disc-only storage engine, dets, had the improvements
described here http://erlang.org/ml-archive/erlang-questions/200605/msg00042.html
. I think this is important because most cheap web apps run in hosted
environments where RAM is scarce and storage capacity is more
important than soft real-time performance. I am actually looking into
running an Erlang app in such an environment (a cheap VPS virtual
server with decent disc capacity but very little RAM), so the dets
issue is definitely on my mind.

Erlang's popularity can also be driven by pure economics. If an Erlang
application server can have x10 the throughput of LAMP server during
high load because Erlang scales much better with large numbers of
parallel processes  (I made up the number but the multiplier shoud be
substantial :) ), web hosting companies could offer Erlang hosting for
lower rates than LAMP hosting (or, vice versa, one would need to rent
fewer dedicated servers to run an Erlang app than a LAMP/Java app).
This would make Erlang more appealing for many web
developers/companies than the alternatives. However, web hosting
companies won't offer Erlang hosting unless developers adopt Erlang in
greater numbers, which depends on features than cater to LAMP
developers, good PR and open source community building.

My 2c.

Yariv