TO:  Selected GCN Sites
RE:  A mistake in the SAX-WFC Notices filtering
DT:  12 Jan 01

I have fixed a mistake in the filtering portion of the GCN system
which incorrectly rejected sending you some of the GCN/SAX_WFC Notices
since June 2000.  I apologize for this mistake.  It is very embarrassing
to make such a mistake.  When I was removing the portions of the
filtering routine that had to do with GRO-BATSE/COMPTEL (GRO was de-orbited
in June 2000), I made a mistake in separating the BATSE stuff from the
SAX_WFC stuff.

You may be wondering how BATSE could be "coupled" to SAX.  Due to limitations
on the packet size (a fixed length since 1993 for backwards compatibility
reasons for the early sites), I had to use a couple of the bits
in the MISC word for different information_content for different SOURCE_TYPES.
In this case, the two sources were BATSE and SAX-WFC.  The filtering looked
at one bit and was supposed to make the correct yes/no_do_you_want_it decision
for the two packet source_types, but I did not code it correctly.

It is too difficult -- nor probably worth it -- to extract from the logfiles
which sites lost which notices, but I can provide a statistical summary.
There were 5 SAX bursts during that time interval.  As of Jan 12
the mistake was effecting 24 out of the 187 sites in the GCN list.  However,
about half of those 24 sites were not part of the GCN list at the June 2000
epoch -- they have been added over the intervening six months.

I am very sorry for this mistake and for the loss of opportunity that this may
have caused you.  I go through a lengthy process of testing each modification
(when new source types, filters, etc are added to GCN), but I messed up
on this BATSE-removal change.

I would like to thank Evert Rol (U of Amsterdam) for noticing that there
were missing GCN/SAX-WFC Notices.

Sincerely,
Scott Barthelmy   (GCN Operations)