Notes from April 16

B. Clark relayed from J. Robnett that security problems with technician
interfaces through HTTP servers are tractable.  HTTP socket connections
are stopped at the router out of NRAO, but a proxy to let in somebody 
from outside, perhaps using a digital certificate as well as password, 
can be implemented in a number of ways.

K. Ryan is seeking information about costs and schedules (for porting to
TC 11 IB) from Esmertec for the Jbed operating system.  He has done additional
experimenting with the demonstration board he has.  For instance, he
made a peek/poke function that runs through telnet, and found he could
send an E-Mail message easily from within the MIB.  C is linked to java
through a mechanism similar to JNI - a declaration is made that makes the
C function look like a java method.

B. Sahr is awaiting price and time quotes on Nucleus+ and ThreadX ports
to the chip.

The last delivery date for the chip development board was April 15.  W. Koski
hasn't seen it.

B. Sahr says that a Nucleus+ simulation that runs under Windows is quite
inexpensive.  He has a demonstration copy.  It requires Microsoft Visual 
Studio, which we do not have installed, but he is intending to do this 
shortly.  Not clear how useful this is in the long run, but may be good for
getting a feel for how it works.

W. Koski is working on defining the properties of the MIB board itself.
He has proposed a standard for the SPI bus, and will see if anybody objects.
Standard he proposes is 1) Clock idle high, 2) Clock low->high clocks data,
3.) MSB first, 4.) 6 MHz clock.  SPI may be converted to balanced pair where
it goes off board, to reduce interference.

G. Hunt feels that the choice of IP stack should be made with great care, 
given the problems they have had with the network implementation in 
GreenBank.

G. van Moorsel reported that the external reviewers for the eVLA software
PDR will be Mark Clark and Ralph Marsen.