[fitsbits] Bintable proposals

[Mike Nolan]

> 
> Assertions that the FITS data format is not suitable for real-time use
> have been made repeatedly over the 22_year history of FITS. 
> 
> I do not agree with these assertions.
>
> ... 
>
> It is important to note that 'using a disk as intermediary' in a
> real-time data acquisition system is a practical and effective solution
> for the NAXIS2 problem in the majority of cases today.

We are using a system that buffers data to disk then writes to 2 DLTs,
and sustains a data rate of 10 MB/s. I like it: The tapes can't quite
keep up, but they always have data available to write, avoiding some
weird behaviors, and the disk slowly fills (over hours).  The data
aren't FITS, but the probably should be. The main users don't keep the
data once they've done their processing, so they're not too concerned
with an exportable format.

However, I don't think it's really reasonable to have your data format
dictate hardware, particularly when the data rates get close to the
edge: if we could sample faster, we would, and that will be true for
another factor of 20 in speed. We're getting to the point where our
desired data rates are comparable to the memory bus speed on a
reasonable computer, and the second pass through the system (disk to
tape) would be a problem.  I think that problem has been true in the
past as well, but not for the past few years.

I am under the impression that Parkes has a similar system with more
samplers (and thus a higher data rate).

Now if we could get somebody to put raw VLBA data into FITS format, that
would help me immensely...

-Mike


---------------------------
Mike Nolan <nolan at naic.edu>
+1 787 878 2612x334  Fax: +1 787 878 1861
Arecibo Observatory / Cornell University
HC03 Box 53995, Arecibo, Puerto Rico 00612  USA