+15Oct02=======================================================================
+1. made some detail changes to the raw USDB data and the routines that
+generate and operate on them. There were some bugs involving a few 'missing'
+callsigns which been fixed. Also there were some, for our purposes, invalid
+callsigns in the database which have been removed (down to about 820,000
+entries now). You should really update your database.
+14Oct02=======================================================================
+1. added show/usdb command as a simple, direct interface to the information
+available in the USDB stuff.
+13Oct02=======================================================================
+1. A large change has occurred. There is now some (optional) US state recog-
+nition. This is in its early stages but appears to work for me. You will need
+to :-
+
+1. download the database at http://www.dxcluster.org/download/usdbraw.gz
+2. save it somewhere (/tmp, anywhere).
+3. Decompress it using your favorite program (gunzip / winzip) [call it
+ usdbraw if winzip]{BE WARNED, some browsers may decompress it on the
+ fly for you, you can tell; if it is 5Mb it is compressed - 16Mb not}.
+4. remove any /spider/data/user.v1 files lying around (at least for this
+ first time.
+5. cd /spider/perl
+6. perl create_usdb.pl /tmp/usdbraw (ie the UNCOMPRESSED file)
+7. Wait, I suggest some cups of tea are in order.
+8. Wait a bit more.
+
+You don't need Compress::Zlib anymore, I assumed it was universal.
+
+You will be able do this while the node is running somewhen. There is a
+planned method of keeping the US DB up to date with smaller (ie < 15Mb) patch
+files once a week but you will have to wait a bit for the code to bed down
+first. You can filter on routes, spots and announces using 'call_state' or
+'by_state' and a comma separated list of state codes. The HELP has been
+updated (which may help...).
+
+Once you have run the create_usdb.pl you will need to restart.
+
+If you don't need this, then don't run create_usdb.pl it will simply be
+a waste of time. The run-time version is 30Mb and has 840,000 odd entries
+in it. This does not replace or supplant sh/qrz (sorry Charlie [who put me
+up to this]).
+
+2. There are a number of other niff-naff changes which I hope will improve
+rather than hinder your user experience (including check forward as well as
+back for those speedy clocked dupes mentioned below).
12Oct02=======================================================================
1. attempt to improve the "check back for 5 minutes to see if this spot is
the same as a previous one but for the time spotted" dupe check.
projects / spider.git / blobdiff