On 18/12/06, Stuart Charlton <[EMAIL PROTECTED]> wrote: > > > > > > > > --- Steve Jones <[EMAIL PROTECTED]> wrote: > > > Which I have.... one of the big problems with IT though is that > > everyone thinks that _they_ are the innovator (Agile "inventing" > > iterations for instance, and that in fact they have discovered the > > silver bullet. > > I've heard you say this before, but I've never heard any of the Agile > leaders claim anything of the sort. It was more of popularizing the > notion that "it's OK to iterate" when one is dealing with uncertainty > and circular dependencies. Something that many have known in other > disciplines, but most project managers didn't accept (and still often > don't!)
I wouldn't say they even popularised it, Spiral and RUP were out well before lots of the Agile approaches. The iteration piece is because I've had lots of "Agile adherents" (maybe acolyte would be a better term) inform me that either a) they are doing agile _just because_ they are doing iterations or b) That I'm doing "agile" because I'm doing iterations. They don't know agile/RUP/Spiral/Waterfall/much at all... but they follow the latest craze like a teenager. You might have spotted that I don't have a very high opinion of a lot of IT :) > > > REST shifts XML docs from A to B, it provides a way for a dynamic > > interface to be communicated to a user. The creates a complex > > programming model that requires significant upfront investment for a > > purported ROI at a "later date". > > I guess we'll have to agree to disagree, as this keeps going in > circles. Understanding REST's value in terms of how it improves "A > talks to B" is not possible, it requires a long view of time, scale, > and control that REST proponents here have been unable to communicate > well and you've been unwilling to concede. Odd that, given the "pay lots for EAI it will make it back in the future" leading to the current EAI as legacy problem, the "J2EE will be easy to upgrade" leading to 6 year old J2EE servers that are as hard to maintain as a mainframe (or harder). REST makes interation between parties more complex, the concept is that this will make things simpler in the long run. Personally I'd bet on something (like HTML and HTTP) which was easier than we had, and which also made it easier in the long run. > > > There is not ONE single thing that > > REST is talking about that hasn't been done previously, and the > > problem sphere of REST is one of the least critical part of any > > successful distributed system (namely the execution context). > > Distributed communication has never been done at the scale of the Web > (scale in terms of diversity of content, heterogeneity of participants, > and marginal cost to integrate). The phone network has more participants than the internet. Sure the content is "only" voice, but its very diverse and very cheap to get on. The global logistics system is pretty huge as well with pretty low cost of entry and with 3PLs and 4PLs is pretty diverse and heterogenous. The web is possibly the "biggest" of these revolutions, but (like Newton) it is built on the shoulders of giants. > > I have little idea what you mean by execution context, as it's a broad > term (and an unfamiliar one). >From the SOA-RM, its basically the bit between consumer and producer that makes things happen, its where WS-* lives and its where REST is. > > > REST is _not_ a silver bullet, remote invocation is _not_ a challenge > > and REST is only disruptive in that it stops people looking for the > > true disruption which will come when we consider remote invocation a > > true commodity. > > In 1971, this might read as: "Relations are _not_ a silver bullet, data > management is _not_ a challenge, and relational databases are only > disruptive in that they stop people looking for the true disruption > which will come when we consider data management a true commodity." Ummm I don't think so, lots of people thought that data was a massive issue in 1971 and the concept of data relations wasn't well understood. Relational databases were disruptive (although not for quite along time) once they became more of a commodity, and the degree to which they were disruptive was that they replaced Hierachical (ummm XML anyone?) data structures. > > I liken REST to relational databases as a disruptive innovation, with > Fielding's thesis akin to Codd's paper in 1970. And yet relations > still generate debate, confusion, and doubt to this day, and those > older systems still exist to a great degree. I'd like REST to XA, its big and complex and lots of the time you can get away without it :) > > Ultimately these debates won't change whether REST is or isn't a > disruptive innovation -- the market determines that. I think in one > sense, it's already proven to be so via the success of web. This is a meme that really has managed to bite. But given that in the first 10 years of the web no-one had heard of REST (pretty much) and no-one was building things using "REST" principles, I think its pretty rich to retroactively claim success. > Integration, on the other hand, does. So it's mostly a matter of the > market shaking out the right complements that are required to take REST > effective for areas where these established approaches rule. Or, from my perspective, the vendors going into yet another technical spiral, doing REST and WS (ala Apache) and there being no actual advance on the real problem which is the definition of the actual business services and the management and governance of business interactions. REST is a "simple" fit for most vendors, whether it has benefits who knows (WS-TX) but it should shift a few more licenses and create a more complex IT infrastructure. I'd really hoped that 2006 would see us move onto the next set of problems. > > Cheers > Stu > > __________________________________________________ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam protection around > http://mail.yahoo.com >
