The question is, is this a bad sign? I do feel these technologies are proven, are moving to mainstream and are not bleeding edge. I feel like these people don't really "get it" and I'm looking to work with and for people who do. - Chris Lasher

when people don't "get it" but they get enough of it that it is a disruptive technology (a small revolution), they're afraid for their position ... hence are even more dismissive of the "it". One aspect of this is that you need to reassure people that their position is not threatened, and technology is their slave, not their master... - General Kafka

Couple thoughts. I'd say the interwebs are pretty good proof that REST works in large projects. RDF I think will always be niche. Ontologies will only be as good as the tools that use them. Regardless, bleeding edge is where all the fun is anyway. :D - Paul J. Davis

Helpful perspectives. Thanks. - Chris Lasher

I would say, use it a prove that it works. - Paulo Nuin

To echo General Kafka's thoughts, if you see 10 new technologies and claim that they'll all fail, you'll be 90% correct. Its just easier to seem smart by being a naysayer. - Benjamin Tseng

Benjamin++ - Egon Willighagen

@Paul- Agreed, REST is widely used, but even though RDF may have a small user base now the use of structured web data formats is important. There are other popular technologies like JSON, XML, databases, and more. Each will get a fraction of this larger overall pie. - Mike Chelen

uhhhh, ontologies haven't proven themselves in a large project?! what world are we in here, exactly? google "gene ontology" for starters - Ian Holmes

Ian, what wet lab research problems have been solved with these methods? That is, I think we struggle here against wet lab high level persons here... - Egon Willighagen

Technologies that may be "proven" for one application may be a poor choice for other applications (think SOAP for doing a financial transaction vs requesting a large FASTA file). It's also important to be aware of the scope of each technology, e.g. is it really smarter to invent your own conventions for identifying and referencing resources on top of the "proven" XML Schema standard, vs using the less proven (though older, btw) RDF standard? - Eric Jain

Eric, you are not comparing SOAP to RDF are you? - Egon Willighagen

It is a very valid lab perspective - you don't use techniques you don't know work. New technology + new result = ambiguity. The error here is in assuming that these techs are experimental techniques, rather than tools for doing stuff. You're not going to "prove" a biological fact with RDF. You're going to manage and analyze data, leading to hypotheses which you will test by verifying their predictions. - Chris Cotsapas

@Egon: Comparing SOAP vs RDF wasn't the intent, but we can do that, too :-) - Eric Jain

Oh, phew... no, let's not do that :) - Egon Willighagen

@Egon: OK scratch the Google search; instead, get your eminent wetlab Luddite to do a *Pubmed* search for "Gene Ontology". 2068 hits. If they plan to have their labrats use microarrays (or any other kind of high-throughput whole-genome assay, for that matter) they'll need to use computable genome annotations in order to make broad inferences about what's up- or down-regulated. And that means using a controlled vocabulary, like GO. Everyone does! Ontologies are not bleeding-edge, they're a fact of life. - Ian Holmes

Ian, yes, I've seen plenty of those papers... it always seemed to me that ontologies are used in microarray analysis, because the data is so dirty that identifying signification up and down regulation based on the data alone (set aside the obvious genes) does not result in anything... GO then helps to reduce this error by averaging regulation for genes in the same GO branch... but really, it's microarray in action... - Egon Willighagen

I was actually more hoping for an example to benefit of ontologies to the biologist that goes beyond fixing experimental uncertainties... where did the use of only ontologies give new scientific insight? Or, what has reasoning provided us so far? - Egon Willighagen

@Egon I kinda liked this approach to using GO: http://tiny.cc/JgUBO Its lacking implementation wise, but the idea is sound. Something I'm planning on revisiting in the future. - Paul J. Davis

GO isn't descriptive enough to be used for automated reasoning on its own (though you can of course do statistical inferences). So if that's your criteria for what is and what isn't an ontology, GO wouldn't be the best example. In any case ontologies (in the wider sense) make automated annotation a lot easier. So perhaps we can rephrase Egon's question to "what new (valid) scientific insights has the extra annotation coverage provided us so far"? - Eric Jain