[erlang-questions] Erlang is the best choice for building commercial application servers

Robert Raschke rtrlists@REDACTED
Tue Mar 13 08:35:28 CET 2012


I have slowly come to the realisation, that the whole Enterprise Java world
is the result of business computing slowly moving away from Cobol. I have
never had much dealings with business computing (e.g., banks, ticket
systems, IT governance) until 5 years ago or so. Up until that point, I was
perplexed by the rise of J2EE. It seemed overly complex to my eyes. But up
until that point I worked first in compilers, then in stats/visualisation.
In hindsight, those areas are pretty niche and allow for quite a lot of
freedom in choice of technology.

The world of business computing, with it's heavy use of VM architectures
(not the new ones, mind you, the old IBM ones, CICS regions on mainframes),
huge databases, communication architectures (CORBA), and business oriented
languages (Cobol, PL/I, and, amazingly, APL) appears to be coming to the
tail end of a transition from IBM/Cobol/CORBA to VMware/Java/J2EE. There's
still enormous amounts of stuff in the old tech there, though. So, this
transition, even after 15 odd years, is not yet done.

One way of getting Erlang into a project, is by using it for prototyping.
But if you do this, you must also be willing to throw away the prototype
afterwards and do it the approved way. Prototypes have a nasty way of
becoming products, though. So, this is one way in. And some of the
libraries that Erlang supplies (the CORBA and SNMP stuff, for example, and
third party things like RabbitMQ) can show applicability in your domain (if
it's not moved fully to J2EE yet).

A rather dramatic trick is to take the risk onto yourself. If you're
reasonably certain that your solution (or prototype :-) will be successful,
and you're willing to take the chance of failing at what you're about to
do, then you always have the option of "just doing it". If it fails and
you're found out to have been using the "wrong" technology, then you're the
culprit. The risk was with yourself. There's nothing that focuses your
attention more than getting yourself into a situation, where "failure is
not an option" :-D This is obviously only possible on a smallish project
with just you and maybe one or two other likeminded individuals.

Robby
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://erlang.org/pipermail/erlang-questions/attachments/20120313/83968040/attachment.htm>


More information about the erlang-questions mailing list