very large data files

[Jason Harris]

Hi,

On Fri, Jul 11, 2008 at 2:11 AM, Eduardo Robles Elvira <edulix at gmail.com> wrote:
> I'm not an admin of GHNS so I don't know what is acceptable there. But if you
> use a cache then when you reposition to somewhere you already have been it
> would read the stars file from your hard disk. Also I think you could have
> both posibilities: downloading all the starts from somewhere (I also think
> that it might be a good idea to download it from NOMAD), and also downloading
> on-the-go.
>

I want to clear something up here:  it is not currently possible to
download NOMAD from anywhere.  USNO distributes NOMAD by having people
mail them a 100 GB USB hard drive, which they will fill with NOMAD and
mail back.  They are sending me the brightest 10% of NOMAD on several
DVDs.

The link Ralf posted lets you download a small chunk of NOMAD data for
a particular region of the sky.  So we could hack this webservice to
stream local chunks of NOMAD into kstars, but the raw NOMAD data would
have to be processed on the fly, and this would be slow as a dog.
Part of what makes handling a huge catalog of stars is that we store
the data in a binary format.  This advantage would be out the window
if we streamed data directly from USNO.  Plus I don't think USNO would
be too happy about this use.

Like I said earlier, I could see us doing a hybrid streaming solution
for the very large catalogs, but I would like to make the 1-million
star and possibly the 10-million star catalogs available for direct
download.  Can I host the files at kde.org?

BTW, my previous file size estimates were too large because I hadn't
taken compression into account.  The 1-million star catalog is only
32MB after compression, so figure around 3GB for the 10-million star
catalog.

Jason