Having worked at Bell Labs, I gotta say the culture of innovation was paramount best described by this statement of an old supervisor: "You should do what you feel is the right thing, it's my job to align that to the needs of the business." So you might spend weeks researching highly available protocols and data distribution techniques to get the one feature correct. It should be no surprise then that the old refriderator-looking phone systems pioneered in the late 70s and 80s had a well known 5 9's of reliability.
A public library as a nice place to get a meeting space, ie for a startup, is a great thing. This post highlights to me the need to expand/improve libraries as both physical and digital resources
HATEOAS yes, "Hypermedia API" no -- reducing the distributed state engine of HATEOAS at the same level of RPC-style APIs doesn't adequately convey the purpose of HATEOAS
I never heard of "Hypermedia API" either - sounds like some REST people are lazy, don't leverage things like content negotiation, URI de-referencing, etc and wind up with an HTTP-RPCish monster with a REST name tag on it.
new entry into the market, Stardog from Clark and Parsia, extremely fast and embeddable / scalable RDF database with OWL reasoning built in (they created the Pellet OWL reasoner)
I agree - 2011 going into 2012 is the rise of the semantic web
There's also Stardog (stardog.com), from the makers of the Pellet OWL reasoner, which just entered this space. AllegoGraph is pretty cool, one of the best examples of commercial LISP success, and they have a pretty rich ecosystem around AllegroGraph (Jena, Sesame, etc support).
Polyglot should not be confused with generalists. There are many generalist programmers who drift through their careers, never becoming an expert in a field so they're easily swapped in and out, easily outsourced, and easily forgotten. Some of the specialists out there, can make some serious salary (e.g. 200-250k/year for those that can build high speed trading platforms, or 150k+ for the architects of the biggest enterprise software).
That being said, limiting yourself to one language will be equally as hurtful, as you are wearing blinders. Even if you don't know all the languages, learning why they did things (DRY, functional programming, etc) can make you a better programmer in Java or whatever language you are in. As you supplement that, you'll find overall skill increases across the board.
My personal opinion is about 1 a year will do just fine, give you enough room to learn more than a book/sample project worth, see where the language goes in a market, and give you enough time to use it for something meaingful.
I'm sure that's where it's headed. You can already build a private IaaS cloud with VMWare's ESX product line, and since their acquisition of SpringSource, who built CloudFoundry, more and more PaaS style capabilities have been produced. I'm sure it won't be long before we see Cloud Foundry Enterprise of vFabric Platform Cloud or whatever name it will be.
I know several people, as well as myself, who are head-hunted reguarily on linked in, and have seen multiple job offers come across this table. All in all, it's not bad. But more importantly, if you've been a part of a big company and worked across teams, then keeping track of all those aquaintances is valuable - they might be your 'in' to finding a job again some day. In that regard, it is near priceless.. I don't want the other software devs on my facebook, but I do want to keep in touch.
Also you can switch to JPA and use any JPA provider