Cone Search

Roy Williams roy at
Thu Apr 25 08:43:12 PDT 2002

Clive, Francois

> I think that the ASU specification (available from
> ) is ahead of you, indeed almost
> six years ahead of you, since it's dated 19-Sept-1996. See the -out.exists
> item.

> I guess it's nice to see new wheels being invented, as long as the
> inventors are aware of the old wheels already in use, otherwise there may
> not be much improvement.

ASU was very much in our minds as we considered specifying the cone search.
For me, I first read the paper in 1997. But ASU gives the *client* a lot of
freedom -- different coordinate systems, SQL queries, Julian dates etc.
Therefore it would take a lot longer for the *publisher* to build an ASU
server from zero than building a cone search server from zero.

The point of the cone search prototype is to see how many data publishers we
can get on the same page, not the sophistication of the interface that they
implement. There have been several VO prototypes for distributed data
publishing, but it seems that if the step is higher than 2 cm, nobody except
the developer seems to get involved. So we experiment with making the step 1

So far we have 9 people in the registry. That is the "new wheels".


Caltech Center for Advanced Computing Research
mail    roy at
voice   +1 626 395 3670

More information about the votable mailing list