[Dx4win] Prefixes/Zones for Spotters bug?

[Jim Reisert AD1C]

This definitely looks like a bug in DX4WIN.  I can duplicate it here also.

I also recommend using the Cluster User program from http://www.ve7cc.net

It will give you find control over the spots you want forwarded to DX4WIN. Plus it does auto-reconnect, which DX4WIN doesn't do very well.

73 - Jim AD1C

 --
Jim Reisert AD1C/Ø, <jjreisert at alum.mit.edu>, http://www.ad1c.us



----- Original Message ----
From: Frank R. Oppedijk <f.oppedijk at avista.nl>
To: Dx4win at mailman.qth.net
Sent: Tuesday, October 21, 2008 5:18:55 AM
Subject: [Dx4win] Prefixes/Zones for Spotters bug?

Hi,

I'm using DX4WIN 7.05.04 and I'm trying to get filtering on prefixes 
to work. Either I totally misunderstand the way the "Prefixes/Zones 
for Spotters" field works, or have discovered a bug.

When I enter OK into the "Prefixes/Zones for Spotters" field, I would 
except to receive only spots from spotters in the Czech Republic. 
However, I get spots from both Czech and US spotters. It looks as if 
DX4WIN interprets the OK as any spots from spotters OK as well as 
spots from spotters from substrings from OK (except the first 
substring - see further!)!

I did a bit more experimenting (spots from I, K, DL, VK, PA, UA, EA 
were available at the DX-cluster at that time):

specified filter prefix spotters displayed from
OK                      OK, K
OI                      I
KI                      I
IK                      K
ODL                     DL
OD                      none (there weren't any OD spots present)
VK                      VK and K
DLVK                    VK and K
PAVK                    VK and K
VKPA                    PA

So it looks like DX4WIN does the following:

1. It chops the specified filter prefix into chunks (eg. the prefix 
VK is split into V and K)
2. It checks which chunks are valid prefixes (V isn't a valid prefix but K is)
3. It matches the spotters against the whole specified filter prefix 
(VK) and all chunks EXCEPT THE FIRST ONE (in this example, only the K 
chunck is checked)

The fact that the first chunk is discarded is not visible from the VK 
example, but is visible from the KI example (K spotters were 
discarded), and similarly from the IK example (I spotters were 
discarded), and also from the DLVK, PAVK and VKPA examples.

Now, have I killed too many brain cells in last weekend's party, or 
have I discovered quite an intricate pattern-matching bug?

As you may understand, the current behavior is undesired for me, as I 
don't have much use for spots from stateside spotters, still get them 
all when specifying a filter containing a K, such as OK...

73,

Frank PA4N