sockets

[Ken Cole]

> It appears that you believe that the license count is how many sockets you
> can have open at any given time, which is not the case.  It is how many
> concurrent processes can use sockets, which is not the same thing.

To be honest, I have to agree with Ed on this one.  If there is a
socketclose() function and we all know there is, then why doesn't it
release the license?  Sockets are not in use anymore by that filePro
process.  If it wants to open sockets again, it will have to fight for
licenses with everyone else.

In all my discussions with fp staff on sockets, and I have had many,
right back to the convention in Philadelphia, I was always led to
believe socket licenses were dynamic based on socket open and close
commands and not on other filePro processes.

We have a sockets license and do not use it for this exact reason.  We
would have bought more as we only got a basic count to do some
testing.  The testing proved we couldn't use fp sockets as the
licensing stood so went another way.

A socket license is nothing like a fp license and should not be
treated that way.  Another great marketing mistake I believe.

Sockets was one thing that could have kept existing fp developers in
fp and maybe even attracted one or two, but not under these license
conditions.

Regards

Ken (the other Ken) Cole