Hi!

I have to write an iphone application that controls a device. This device exposes a telnet based interface. The application should ideally have user access control and customizability for each user.

I was thinking of writing C++ classes that would communicate with the device using sockets. This functionality can then be exposed through web-services that can be called by the iphone application.

However as i looked into it deeper, the api allows you to register for events using telnet and then you can receive notification when those events occur. That kinda put a spanner in the works for me. I for one dont know a "push" scenario can work with webservices.

First off i have never programmed for the iphone so far. So i am not really sure what can be done. So i was thinking if instead of having a webserver to go through, why not have the application independently running on the iphone, directly communicating with the device using sockets. The question though is, is that possible and second i am thinking it would raise a security aspect. First we could control security as everything was going through our central server. Is there a way to handle security (in the sense who has access to the device) without having a central server.

I am sorry that this seems like an unorganized post, but iam trying to brainstorm here.

Looking forward to hear your opinions.