1. Mate has stacks at each three contexts and a stack is isolated for return-address and operand.
I'm wondering about reason to divide stack into return-addr and operand.
I'm guessing it is due to security/robustness. but I think it has more flexibility with one-stack than with isolated stack. for example, more calling depth with less operands, or more operands with less calling depth.
Two stack implementations are rather simple, more simple than single stack. The initial versions of Mate were very much influenced by FORTH, which uses this model. I chose two stacks because it seemed the minimalist way to achieve the goal of having subroutines.
Your statement about sharing is correct; however, it is contrary to the general allocation policy TinyOS uses. Shared resources mean that it is harder to reason about when the resource run out. If, instead, each piece of functionality (e.g., component) statically allocates what it needs, then it can control the usage of the resource in the absence of any interference and interactions.
2. It seems that anybody can install new network application into sensor field. but I looks very dangerous. Is there any mechanism for security? for example, key for installation... etc.
The VM itself, in the ASPLOS incarnation, had no measures for security. Using TinySec would provide code integrity and confidentiality. However, a few security issues remain in that situation, notably cut-and-paste attacks.
There are solutions to most of the security issues the VM faces, but at this point they are primarily academic. That is, were the VM to be deployed somewhere such as Great Duck Island, it's unlikely that there are adversaries who want to take control of the network.
3. I met mate virtual machine on a paper 'Mate: A Tiny Virtual Machine for Sensor Networks', and I guess that many development/research is going on about Mate. but, There are few information about Mate and It seems that TinyOS is main development stream in these days.
Let me know further works about Mate virtual machine after the paper I read and future of Mate virtual machine.
It is something I am working on right now. There should be developments in a few months. Specifically, users will be able to easily build customized VMs, and there will be scripting languages that compile down to VM bytecodes.
Phil
-------
"We shall not cease from exploration And the end of all our exploring Will be to arrive where we started And know the place for the first time."
- T. S. Eliot, 'Little Gidding'
_______________________________________________ Tinyos-users mailing list [EMAIL PROTECTED] http://mail.Millennium.Berkeley.EDU/mailman/listinfo/tinyos-users
