Basically, the problem is that you need to build a robust implementation of the protocol in MUMPS that is capable of handling the same load as your dedicated box. In HTTP terms, you;re just building a tunnel that doesn't really add any value to what you already have. But MUMPS implementations typically don't give you the tools you need to write a proper server. Maybe Cache' is an exception (I don't know), but although it is possible to passively connect, running netstat shows me that MUMPS based servers don't really "listen". What is badly needed is something like the listen() and accept() calls in BSD sockets. Worse, there is no way (that I can think of at least) to either write a multi-threaded server or to do a select(), allowing for the possibility of asynchronous I/O. Finally, it is not clear that it is possible for one process to start reading data from a stream while the network connection is still open and another process is copying data to that stream. This forces the server to simply wait until a complete message has been recerived (and probably copied to a global) and only then can processing begin.
--- Renee Cannon <[EMAIL PROTECTED]> wrote: > Why is your solution less ridiculous than the one you are ridiculing? > Your > solution seems more complex in a situation where the components > needed to > solve the problem are already in place. > > I'm just curious- I love architecture discussions. > > Thanks for indulging me, > Renee > > -----Original Message----- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of > Greg > Woodhouse > Sent: Monday, May 02, 2005 12:13 PM > To: hardhats-members@lists.sourceforge.net > Subject: RE: [Hardhats-members] WebServices future development??? > > In fact, my inclination would be to not try to implement web services > directly in MUMPS, but there still needs to be a mechanism for > communicating with whatever you do use to implement web services. > Using > the TCP/IP as natively implemented in MUMPS isn't a very good > solution. > I would much prefer to see a FIFO or some other mechanism for > invoking > a C library capable of using JMS (say) to talk to the web service > implementation. But using HTTP implementedf in MUMPS to talk to > another > box (or process) implementing wb services using HTTP is a little > ridiculous. > > > > --- Dan <[EMAIL PROTECTED]> wrote: > > Web service was basically introduced as a standard to replace > CORBA. > > > > There's a lot more to it than just being XML. It's a way to expose > > > procedures to 3rd parties without having to develop your own set of > > > protocols. You expose your service via a WSDL (web service > > descriptor > > language) document. This lets the client know how to connect to > your > > > > service and what to expect in return. There are two types of > > services -- > > document and rpc. > > > > Most languages have the necessary components to build servers and > > clients > > from blocks of code that follow certain conventions. > > > > I'm not sure if VistA would benefit from WS being implemented > > natively. I > > don't think the VA is pursuing XML development with MUMPS (although > > very > > well suited to handle XML). I think they've stopped development on > > their > > XML parser. > > > > It would probably be easier to expose services through a supported > > implementation, such as Tomcat/Axis and then connect to M for > > database access. > > > > > A practical man is a man who practices the errors of his forefathers. > --Benjamin Disraeli > ==== > Greg Woodhouse > [EMAIL PROTECTED] > [EMAIL PROTECTED] > > > > > > ------------------------------------------------------- > This SF.Net email is sponsored by: NEC IT Guy Games. > Get your fingers limbered up and give it your best shot. 4 great > events, 4 > opportunities to win big! Highest score wins.NEC IT Guy Games. Play > to > win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=20 > _______________________________________________ > Hardhats-members mailing list > Hardhats-members@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/hardhats-members > > > > ------------------------------------------------------- > This SF.Net email is sponsored by: NEC IT Guy Games. > Get your fingers limbered up and give it your best shot. 4 great > events, 4 > opportunities to win big! Highest score wins.NEC IT Guy Games. Play > to > win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=20 > _______________________________________________ > Hardhats-members mailing list > Hardhats-members@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/hardhats-members > A practical man is a man who practices the errors of his forefathers. --Benjamin Disraeli ==== Greg Woodhouse [EMAIL PROTECTED] [EMAIL PROTECTED] ------------------------------------------------------- This SF.Net email is sponsored by: NEC IT Guy Games. Get your fingers limbered up and give it your best shot. 4 great events, 4 opportunities to win big! Highest score wins.NEC IT Guy Games. Play to win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=20 _______________________________________________ Hardhats-members mailing list Hardhats-members@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hardhats-members
