This was a draft of version 4 the catalogue of nearby stars.  This was
started May 2007 in cooperation with Hartmut Jahreiss, ARI (that's
cns4.fin).

Then, about 2017, Christian Dettbarn gave us his state of the thing
as cns-5.2, and we updated the the RD a bit.

In November 2018, We got another version of the CNS from Jahrreiß
(cns4_vanl), while at LSW, Alexander Golovin started a project to polish
the thing.

cns4_vanl still had plenty of quirks, too many actually to make
publication make sense.  Hence, we never published cns4.  We're just
keeping this around in case later historians want a peek.


Quirks
======

The sun is in the catalogue but has no id.  We can't have that and thus
we skip it.  That is why the grammar gives *two* header lines.



======
Update CNS 4

(1) Can we have a reliable documentation on how to read the "No" column?
I know about the first column, but it'd be useful to know what the
policy on the component designations is, and I have little idea what to
do with the last few columns (h, n, sn,...) -- well, col. 8 has to do
with binaries recognised by Hipparcos, but there's characters in there
(xChns)?  And col. 9 has something to do with erroneous inclusions, and
again there's odd chars in there (xnpqs)...

(2) I'm just mapping thorugh YPC -- not quite sure what it is.
An oddly formed YPC number?  But what's the purpose of the extra 00 and
10?  What's HIL in there?

(3) Having an official guide for column 174, the code for the
Durchmusterung source, would be good if we wanted to keep it (I'd lean
towards tossing it, though I could imagine quite a few original sources
might have used BD identifiers...)

(4) Dettbarn had cross-identifications into 2MASS and Simbad -- no idea
how well curated they were.  I suspect doing the cross matches now and
with TAP available gets you better data with less effort, but I can't
say for sure.  I *suppose* we could relatively easily recover
Christian's matches, though.

(5) I can't really guess what the "D" column might be (col 186-190).
I'm now carrying it through as id_d.

(6) To save you a bit of work, in addition to pm and pmpa I'm also
giving conventional pmra and pmdec.  Here's some ADQL that spits out the
records where pmdec is badly off vs. Hipparcos:

select a.*, abs(a.pmdec-b.pmde)/b.pmde as pmdec_dev
from cns4.data as a
join hipparcos.main as b
  on (b.hipno=a.id_hip)
where abs(a.pmdec-b.pmde)/b.pmde>0.1
order by pmdec_dev desc

(looking at the histogram I'd say there's very few that are obviously
wrong and need some human eye).  This needs improvement to properly
treat multiple stars (with the hipparcos.main column whatPos, I suppose).

Oh, don't worry about the errors TOPCAT flags in that query: That's
again because I'm hiding the CNS table metadata from TOPCAT to keep
people from accidentally stumbling on it.

(7) the pi_code (cols 109-113) apparently has a lot of new characters,
so the footnote is out of date.  I suppose it's more efficient to figure
this out for the few objects that don't have either Gaia or Hipparcos
parallaxes than to work this out.

(8) A first step to obtain good astrometry I'd take is to use the
objects that have a Hipparcos counterpart and those that have a
counterpart in HD, which can be halfway reliably matched with Gaia using
the hdgaia catalog also available in the data center.  For now, I'd
bring everything to epoch 2000 using the ivo_apply_pm user defined
function (see the service tab in TOPCAT's TAP window).  Here's a query
that adds the respective data where available; arrays in TOPCAT work
like they do in Java or Python, i.e., just use [0] and [1] to get to the
components:

select
  cns.*,
  ivo_apply_pm(g.ra, g.dec, g.pmra/3.6e6, g.pmdec/3.6e6, -15) as gaiapos,
  g.source_id as gaia_id,
  ivo_apply_pm(h.ra, h.dec, h.pmra, h.pmde, 8.75) as hippos
from
  cns4.data as cns
  left outer join hipparcos.main as h on (h.hipno=cns.id_hip)
  left outer join hdgaia.main as hg on (hg.hd=cns.id_hd)
  left outer join gaia.dr2light as g on (hg.source_id=g.source_id)

See http://docs.g-vo.org/adql/html/page012.html on why there's a "left
outer" here.  Making a plane plot of gaiapos[0]-hippos[0] against
gaiapos[1]-hippos[1] suggests that there are surprisingly few actual
problematic objects here.  Also note that when you open Aladin next to
Topcat, you can tell TOPCAT in Views/Activation Actions to "Send Sky
Coordinates" (check the box).  You'll probably want to configure
hippos[0] and hippos[1] as the columns to send.


(9) Two other tables I'd recommend looking into are arihip.main, which
is a careful rereduction of Hipparcos that's useful when motions are
non-linear (which may be quite common in this sample) and gdr2dist.main
that's a sane start for using Gaia parallaxes (and praise the lord if
you can get away with it rather than have to deal with the Gaia
parallaxes and their proper interpretation yourself...).