An interesting posting from Sam here. He asserts, for next gen web architectures, the following are the long terms bets (with my commentary, in italics):
- REST (yes)
- Hadoop (yes. Map/Reduce (which to my biased database eyes is just "group by on f(x)" as executed in parallel databases everywhere, but if it becomes popular outside the google/yahoo halls, then it is likely to be transformative))
- Erlang (not an expert)
- Jabber (not an expert)
- Microformats (very interesting, is it gaining the right amount of traction, if so, yes...)
Let me comment on one aspect of Sam's posting -- the assertion that all data are trending to be in one place (SaaS model for data) and how it affects the future of traditional database/SQL centric architectures... I think the impact within enterprises of "wouldn't it be nice if all data were in one place?" vision to be very very slowly realized. However, if we think about whether traditional ways of information access are getting supplanted, that I grant. See the point below...
Now, of course, I have to add to Sam's list, I would not be much of a blogger if I did not, would I? If one of 3 or 4 get eliminated (how many more programming languages, however pure, does the world need, Sam?), then I think "light semantics" makes the cut. Why? The OWL/RDF approach is too much, too late, for most applications. However, it is very difficult to make sense of information without some semantics. To me, as I have said in an earlier post, without some semantics, there is no information access, and that, to me, is a far bigger transformation, at lease the way I see it, than SaaS, for my customers...
Anyway, thanks Sam.. Always a fun talking to you and reading your blogs.
Comments