Re: [nvo-techwg] Time series data

From: Rob Seaman <seaman-at-noao.edu>
Date: Tue, 26 Jun 2007 16:04:01 -0700


On Jun 26, 2007, at 3:39 PM, Arnold Rots wrote:

> I only asserted that STC *can* provide the functionality.

It was the combination with Doug's statement that VO needs a single common time series DM that triggered my response. In such a case, either STC is THE solution, or it is not A solution - whatever you may assert.

> And, of course, it does that with minimal change to the VOEvent
> standard and schema - that's leveraging capabilities that are
> already built in.

Which is precisely why I'll be surprised if we don't adopt STC for representing orbital elements. My only concern there is to reach a consensus that hyperbolic orbits and other important edge cases are properly supported. In short, can STC represent canonical data like MPC, etc?

> Anyway, 437 bytes is peanuts compared to the 18 kB of your message ;-)

I saved the message as txt - 8451 bytes with a full header, 5964 with a minimal header. Smiley noted, but would this be an appropriate overhead for every point of an extensive light curve? If you are indeed seeing a transport overhead of a factor of 2 or 3 (two byte unicode, perhaps?), this rather strengthens my argument.

Rob Received on 2007-06-27Z00:34:58