[
Lists Home |
Date Index |
Thread Index
]
On Friday 01 March 2002 02:54 pm, Paul Prescod wrote:
> First, you've taken one argument and acted as if it is many. The
> vast majority of the arguments for REST have nothing to do with
> programmer competency and everything to do with making systems that
> interoperate at scale versus in labs.
Interoperability at scale is a red herring. There are vast systems
that interoperate that don't use REST.
A large part of the trend we see is toward very loosely coupled/typed
systems and robustness in the face of evolution.... kind of the same
thing that increasingly pushes us toward scripting languages vs.
assembler. REST plays well here which is a large part of it's value.
> Second, it is very common in the security world to promote systems
> that promote security, because no system can in and of itself
> guarantee security. I see nothing wrong with choosing an
> architecture because it might tend to lower the number of security
> holes.
Sure, but we haven't seen anything in REST that would truly reduce the
number of holes.
|
