The PPMXL catalogue.  There's a publication on this, see there.

This stuff was generated from a dump of the obspos data (see usnob) by
S. Röser using some Fortran code available in src.

To generate the final table, some SQL scripting in q.rd was used to

* add 2MASS magnitudes
* merge some PPMX stars.

Merging 2MASS mags was excruciatingly slow using both a straight
join-update and a cursor-based solution (that's the import_rawdata
data).



Fixes
=====

We removed the bad data listed in Barron et al (2008).  Plus, the
following ipixes were silently removed after the fact (all spurious
detections in, e.g., galaxies):

  1806716380370098071
  1806716381171609887
  1806716381403930109
  1806716386911328060
  1806716392926462929
  1806716398727130955
  1806716393011495728
  1806716398755758034
  1806716398979337998
  1806716402243365626
  1806716403570106141
  1806716403551888880
  1806716403904907450
  1806716403953034727
  1806717137430219103
  1806716404947521035
  1806716405006121268
  1806716405394628940
  1806716405439940645
  1806716406193244908
  1806716406203320646
  1806716404833588364


A bug was discovered in PPMX (no L!) where Ru magnitudes were computed
from 2MASS magnitudes even when 2MASS magnitudes contained null values.
This is fixed in PPMX, but since we didn't want to repeat the join,
the (silent) fix was done in-database.  This is *not* necessary on
a re-import, since PPMX itself has been fixed.  The query ran was

update ppmxl.main
set r1mag=null
where
	flags&2=2
	and r1mag is not null
	and  (kmag is null or jmag is null or hmag is null)

This touched 6169 rows.



John's PM corrections
=====================

In 2014, John Vickers established a better rest system at least for
2MASS stars fainter than 14th magnitude.  We use a python procedure
to add the vickers_pmra and vickers_pmde columns.  To make this work,
in Debian you need to install the postgresql-plpython-<version> package.

To extend an existing PPMXL table without re-importing everything,
say

	alter table ppmxl.main add column vickers_pmra real;
	alter table ppmxl.main add column vickers_pmde real;

and then

sh johnspms/add_vickers_pms.sh

The original code by John is in johnspms.

The accompanying publication is 2016AJ....151...99V


The Static Dump
===============

People want to pull this stuff whole.  To make this easier, we provide an
ASCII dump in static/dump-for-web.gz.  It is generated my bin/mkasc.py;
see that file for details.

Since we don't want crawlers to pull this stuff directly, and we offload the
actual download to an apache, there's bin/download.cgi and

	<Directory /data/gavo/inputs/ppmxl/>
		Require all granted
	</Directory>

	ScriptAlias /ppmxl /data/gavo/inputs/ppmxl/bin/download.cgi

in /etc/apache/sites-enabled/000-default.conf