In message news:9r7oui$18ve$1@agate.berkeley.edu, Eric J. Korpela wrote:

Now that I'm aware that I can talk about it, I suppose I could fill everyone in on the current direction things are going (colored a bit with the direction I'd like them to go).

The first big thing is that observing at Parkes will be different than at Arecibo.  We'll be tagging along on the primary receivers, so we won't be slewing across the sky like we do at Arecibo.  Gaussian fitting goes away in this observing mode.  On the other hand we're on the same spot of sky for longer, so we can bump up max FFT length to 256k and look at even narrower bandwidths.

The current plan is for a workunit to still be 1M samples, but to frequency step across the workunit.  The first 256K samples will be at in one frequency band.  The next 256K samples will be in a band 2.5 MHz higher.  So in essence each workunit will cover about 40 kHz rather than the current 10 kHz.  Going to 256K will increase processing time an as yet undetermined amount.

I'm pushing for the new client to be capable of processing Arecibo work units, so I'm hoping that we'll still keep looking in the north without needing some people to keep running the old version of the client.

We're still nailing down the details.  And we haven't even started client development at this point.  (We're pretty strapped for cash right now, too.
A lot of our funding sources are feeling the economic pinch and some donations that were to be matched by the university haven't (yet) come through.  So if anyone has some loose change...)

Of course, we send our thanks to those who have given cpu time, advice, money, equipment, etc.