--- In [email protected], Gregg Wonderly <[EMAIL PROTECTED]> wrote: > > Alexander Johannesen wrote: > > I guess it depends on who you are and what you're trying to do. But > > isn't it a fair assumption that if you are to be successful with > > networks across distances (and a lot of SOA falls into this category), > > the web is unavoidable unless you've got too much money to burn? > > I think it depends on what you are doing. I use Jini across long haul networks, > throught the internet and through VPNs all the time. My desktop client > environment contacts lookup services, downloads code (yes via HTTP), fires up > the services GUI (through the serviceUI mechanisms) and then proceeds to carry > out the whole user session with nothing but JERI. HTTP plays a small part here. > In fact, I use a protocol type, vhttp:, to download code so that I can "cache" > and expose the downloading for/to the user so they know what they are waiting > for. The handler for that protocol type can defer to any underlying transport > via the Java plugable protocol handler stack. > > So, http plays an extremely small, and not necessary part of all of my systems. > > Gregg Wonderly
You may all recall that I have raised the question of possible Alternatives to WS Standards for implementing a SOA a couple of times. The main ones proffered so far have been REST, people and flying monkeys. Gregg, this is the closest that any one has come to positing those Orphan Java technologies, Jini and JERI (I wonder, BTW, what Walt Disney could do with these names??) as an alternative. Thank you, Gregg. Gervas >
