|
SpotCollector's?prefiltering feature is very helpful.?
One thing that I would appreciate is an additional filter, blocking the spots of? clearly unreasonable callsigns produced by digital mode skimmers.
Recently I have seen calls like 5WGYH86QK6J, EZZE3K8A3LL, and now a few? minutes ago, STZJ4GCNWEE spotted on digital modes. When the last one? appeared I got an alert since I have yet to work Sudan on 40m.
There are some quite strange special event callsigns, such as OU0POLIO and DL2025B,? so the filter cannot be too strict but it would be helpful if the most obvious could be blocked. As a user option, of course.
Some ideas for rules would be?
Base callsigns (with any prefix or suffix stripped) starting with a letter having more than one group of digits Base callsigns starting with a digit that is followed by more than two letters or having more than two groups of digits? Base callsigns having more than two letters before the first digit
Bj?rn SM7IUN
|
+ AA6YQ comments below
SpotCollector's prefiltering feature is very helpful.
One thing that I would appreciate is an additional filter, blocking the spots of clearly unreasonable callsigns produced by digital mode skimmers.
Recently I have seen calls like 5WGYH86QK6J, EZZE3K8A3LL, and now a few minutes ago, STZJ4GCNWEE spotted on digital modes. When the last one appeared I got an alert since I have yet to work Sudan on 40m.
There are some quite strange special event callsigns, such as OU0POLIO and DL2025B, so the filter cannot be too strict but it would be helpful if the most obvious could be blocked.
As a user option, of course.
Some ideas for rules would be
Base callsigns (with any prefix or suffix stripped) starting with a letter having more than one group of digits Base callsigns starting with a digit that is followed by more than two letters or having more than two groups of digits Base callsigns having more than two letters before the first digit
+ I've sent you a new version of SpotCollector that provides a new "Discard spots of invalid callsigns" option on the Pre-filtering window that when enabled ignores spots whose base callsigns
- start with a letter having more than one group of digits
- start with a digit that is followed by more than two letters
- have more than two groups of digits
- have more than two letters before the first digit
+ Please let me know how it goes...
73,
Dave, AA6YQ
|
For all AR clusters you can enter a filter command that takes advantage of spot quality filter: SET DX FILTER NOT SKIMBUSTED
?
That does a great job of eliminating bad RBN/skimmer spots, based on work by CT1BOH. Not perfect, but really big step forward.
?
73 John K3TN
?
?
|
Try this: SET DX FILTER NOT
SKIMBUSTED AND UNIQUE=3
Big change in how it reacts!! Definitely lower busted calls.
Will WC2L
On 03-Feb-25 7:41 AM, John K3TN via
groups.io wrote:
For all AR clusters you can enter a filter command that takes
advantage of spot quality filter: SET
DX FILTER NOT SKIMBUSTED
?
That
does a great job of eliminating bad RBN/skimmer spots, based
on work by CT1BOH. Not perfect, but really big step forward.
?
73
John K3TN
?
?
--
William Liporace WC2L
or
AR-Cluster Node telnet dxc.wc2l.com 7373 or 144.93 MHz
wc2l@...
|
For DXSpider clusters one can set a rbn reject filter with q:2
(i.e., rbn/skimmer spots must be q:3 or higher)
Between "not skimbusted" on AR Cluster nodes and "reject q:2"
on DXSpider nodes, I do not ever recall ever seeing any of the
"unreasonable" callsigns but I do see the really "wild" special
event calls since they are spotted by multiple skimmers (and
thus confirmed within the RBN system).
73,
... Joe, W4TV
toggle quoted message
Show quoted text
On 2025-02-03 7:41 AM, John K3TN via groups.io wrote: For all AR clusters you can enter a filter command that takes advantage of spot quality filter: SET DX FILTER NOT SKIMBUSTED
That does a great job of eliminating bad RBN/skimmer spots, based on work by CT1BOH. Not perfect, but really big step forward.
73 John K3TN
|
Is JH1RFM not working?
Steve K2SN
Get
On Feb 3, 2025, at 9:53 AM, "Joe Subich, W4TV" < lists@...> wrote:
For DXSpider clusters one can set a rbn reject filter with q:2
(i.e., rbn/skimmer spots must be q:3 or higher)
Between "not skimbusted" on AR Cluster nodes and "reject q:2"
on DXSpider nodes, I do not ever recall ever seeing any of the
"unreasonable" callsigns but I do see the really "wild" special
event calls since they are spotted by multiple skimmers (and
thus confirmed within the RBN system).
73,
... Joe, W4TV
On 2025-02-03 7:41 AM, John K3TN via wrote:
For all AR clusters you can enter a filter command that takes advantage of spot quality filter: SET DX FILTER NOT SKIMBUSTED
That does a great job of eliminating bad RBN/skimmer spots, based on work by CT1BOH. Not perfect, but really big step forward.
73 John K3TN
|
The two addresses listed on
dxcluster.info don't appear to be working.
Will WC2L
On 03-Feb-25 10:31 AM, Stephen
Rabinowitz via groups.io wrote:
Is JH1RFM not working?
Steve K2SN
Get
On Feb 3, 2025, at 9:53 AM, "Joe Subich,
W4TV" < lists@...>
wrote:
For DXSpider clusters one can set a rbn reject filter with q:2
(i.e., rbn/skimmer spots must be q:3 or higher)
Between "not skimbusted" on AR Cluster nodes and "reject q:2"
on DXSpider nodes, I do not ever recall ever seeing any of the
"unreasonable" callsigns but I do see the really "wild" special
event calls since they are spotted by multiple skimmers (and
thus confirmed within the RBN system).
73,
... Joe, W4TV
On 2025-02-03 7:41 AM, John K3TN via wrote:
For all AR clusters you can enter a filter command that takes advantage of spot quality filter: SET DX FILTER NOT SKIMBUSTED
That does a great job of eliminating bad RBN/skimmer spots, based on work by CT1BOH. Not perfect, but really big step forward.
73 John K3TN
--
William Liporace WC2L
or
AR-Cluster Node telnet dxc.wc2l.com 7373 or 144.93 MHz
wc2l@...
|
Dave,
? Would KH6/K0VM pass '- start with a letter having more than one group of digits' test ?
AL, K0VM
On 2/3/2025 4:04 AM, Dave AA6YQ wrote:
toggle quoted message
Show quoted text
+ AA6YQ comments below
SpotCollector's prefiltering feature is very helpful.
One thing that I would appreciate is an additional filter, blocking the spots of clearly unreasonable callsigns produced by digital mode skimmers.
Recently I have seen calls like 5WGYH86QK6J, EZZE3K8A3LL, and now a few minutes ago, STZJ4GCNWEE spotted on digital modes. When the last one appeared I got an alert since I have yet to work Sudan on 40m.
There are some quite strange special event callsigns, such as OU0POLIO and DL2025B, so the filter cannot be too strict but it would be helpful if the most obvious could be blocked.
As a user option, of course.
Some ideas for rules would be
Base callsigns (with any prefix or suffix stripped) starting with a letter having more than one group of digits Base callsigns starting with a digit that is followed by more than two letters or having more than two groups of digits Base callsigns having more than two letters before the first digit
+ I've sent you a new version of SpotCollector that provides a new "Discard spots of invalid callsigns" option on the Pre-filtering window that when enabled ignores spots whose base callsigns
- start with a letter having more than one group of digits
- start with a digit that is followed by more than two letters
- have more than two groups of digits
- have more than two letters before the first digit
+ Please let me know how it goes...
73,
Dave, AA6YQ
|
+ AA6YQ comments below
Is JH1RFM not working?
+ That DX Cluster is still down.
73,
Dave, AA6YQ
|
# AA6YQ comments below
Would KH6/K0VM pass '- start with a letter having more than one group of digits' test ?
# Yes: the base callsign is K0VM, which does not have more than one group of digits.
73,
Dave, AA6YQ
+ AA6YQ comments below
SpotCollector's prefiltering feature is very helpful.
One thing that I would appreciate is an additional filter, blocking the spots of clearly unreasonable callsigns produced by digital mode skimmers.
Recently I have seen calls like 5WGYH86QK6J, EZZE3K8A3LL, and now a few minutes ago, STZJ4GCNWEE spotted on digital modes. When the last one appeared I got an alert since I have yet to work Sudan on 40m.
There are some quite strange special event callsigns, such as OU0POLIO and DL2025B, so the filter cannot be too strict but it would be helpful if the most obvious could be blocked.
As a user option, of course.
Some ideas for rules would be
Base callsigns (with any prefix or suffix stripped) starting with a letter having more than one group of digits Base callsigns starting with a digit that is followed by more than two letters or having more than two groups of digits Base callsigns having more than two letters before the first digit
+ I've sent you a new version of SpotCollector that provides a new "Discard spots of invalid callsigns" option on the Pre-filtering window that when enabled ignores spots whose base callsigns
- start with a letter having more than one group of digits
- start with a digit that is followed by more than two letters
- have more than two groups of digits
- have more than two letters before the first digit
+ Please let me know how it goes...
73,
Dave, AA6YQ
|
There are two reasons that I quickly implemented Bj?rn's suggestion:
1. it works ignores invalid callsigns from all spot sources, no matter what flavor of cluster management software they are running
2. if you're using the direct interoperation between DXLab and WSJT-X, it works ignores invalid callsigns decoded by local instances of WSJT-X
???? 73,
??????????? Dave, AA6YQ
|
Whats the difference between, if any, between NOT SKIMBUSTED and SKIMVALID ? My skimmer filters use SKIMVALID and other than occasionally getting a T in front of some calls, they seem pretty reliable …..
toggle quoted message
Show quoted text
On Feb 3, 2025, at 7:41?AM, John K3TN via groups.io <jpescatore@...> wrote:
? For all AR clusters you can enter a filter command that takes advantage of spot quality filter: SET DX FILTER NOT SKIMBUSTED
?
That does a great job of eliminating bad RBN/skimmer spots, based on work by CT1BOH. Not perfect, but really big step forward.
?
73 John K3TN
?
?
|
Dave, how would that treat the infamous P5RS7??
There are anniversary celebrations that do have odd constructions of call signs, some that had years for the numerical digit.?
I don’t go out of my way to work those types of long call signs.?
But, one in a while I’ll go for one of the rare prefixes.
?
73,
Dave, w6de
?
toggle quoted message
Show quoted text
From: [email protected]
<[email protected]> On Behalf Of Jamie WW3S via groups.io
Sent: 3 February, 2025 22:05
To: [email protected]
Subject: Re: [DXLab] Reasonable call prefilter
?
Whats the difference between, if any, between NOT SKIMBUSTED and SKIMVALID ? My skimmer filters use SKIMVALID and other than occasionally getting a T in front of some calls, they seem
pretty reliable …..
On Feb 3, 2025, at 7:41?AM,
John K3TN via groups.io <jpescatore@...> wrote:
?
For all AR clusters you can enter a filter command that takes advantage of spot quality filter:
SET DX FILTER NOT SKIMBUSTED
That does a great job of eliminating bad RBN/skimmer spots, based on work by CT1BOH. Not perfect,
but really big step forward.
|
I posted this on the N1MM Logger+ group just over a month ago:
?
See <http://www.k3lr.com/w9zrx/AR-Cluster%20Filter%20Commands.pdf>, near the top of page 5.
?
Looking at those definitions: Skimvalid looks for multiple spots with the same callsign. Skimbusted looks for similar (e.g. "off-by-one") callsigns that are already marked as valid. A busted spot could still pass as Skimvalid if enough skimmers bust the call the same way, but if the correct call has already been spotted there, adding Not Skimbusted to the filter may help prevent the busted spot from showing up.
?
73,
Rich VE3KI
?
On Mon, Feb 3, 2025 at 05:04 PM, Jamie WW3S wrote:
Whats the difference between, if any, between NOT SKIMBUSTED and SKIMVALID ? My skimmer filters use SKIMVALID and other than occasionally getting a T in front of some calls, they seem pretty reliable …..
?
On Feb 3, 2025, at 7:41?AM, John K3TN via groups.io <jpescatore@...> wrote:
For all AR clusters you can enter a filter command that takes advantage of spot quality filter: SET DX FILTER NOT SKIMBUSTED
?
That does a great job of eliminating bad RBN/skimmer spots, based on work by CT1BOH. Not perfect, but really big step forward.
?
73 John K3TN
?
?
|
The discussion below assumes one is connecting to a node running AR Cluster software written by the late AB5K.? Users need to send filter commands that, in my opinion, are complicated in order
to customize the spots they receive, either to eliminate bad spots or restrict the spots by band, mode, origin, etc.
?
CC Clusters, which use node software by Lee VE7CC, have built-in dupe filters and validation checks, so users don’t have to send a command to do that.
?
The two most popular CC Cluster nodes in North America are VE7CC and AE5E.? The latter is at the HQ of Digi-Key in northern MN, so it has commercial internet service, which makes it more reliable
that nodes using regular, residential internet service.? Another of interest is N2WQ-1, which is hosted by AWS, Amazon Web Services, so it’s reliable and fast.
?
If you do want to filter at the node, which I especially recommend for contesting with a logging program, such as N1MM, you can use a free program CC-User, written by VE7CC, to set up different
filters for contests and regular DXing.? Info at:
?
Jim N7US
?
toggle quoted message
Show quoted text
From: [email protected] <[email protected]>
On Behalf Of ve3ki via groups.io
Sent: Monday, February 3, 2025 5:22 PM
To: [email protected]
Subject: Re: [DXLab] Reasonable call prefilter
?
I posted this on the N1MM Logger+ group just over a month ago:
See <>, near the
top of page 5.
Looking at those definitions: Skimvalid looks for multiple spots with the same callsign. Skimbusted looks for similar (e.g. "off-by-one") callsigns that
are already marked as valid. A busted spot could still pass as Skimvalid if enough skimmers bust the call the same way, but if the correct call has already been spotted there, adding Not Skimbusted to the filter may help prevent the busted spot from showing
up.
On Mon, Feb 3, 2025 at 05:04 PM, Jamie WW3S wrote:
Whats the difference between, if any, between NOT SKIMBUSTED and SKIMVALID ? My skimmer filters use SKIMVALID and other than occasionally getting a T in front of some calls, they seem pretty reliable …..
?
On Feb 3, 2025, at 7:41?AM, John K3TN via groups.io <jpescatore@...> wrote:
For all AR clusters you can enter a filter command that takes advantage of spot quality filter:
SET DX FILTER NOT SKIMBUSTED
That does a great job of eliminating bad RBN/skimmer spots, based on work by CT1BOH. Not perfect, but really big step forward.
|
+ AA6YQ comments below
Dave, how would that treat the infamous P5RS7?
+ That would be ignored if you enabled the Pre-filtering window's new "ignore invalid callsigns" option - which had it been available back in 1992-1993 would have saved us all a lot of time. ? ? ? 73, ? ? ? ? ? ? Dave, AA6YQ
|
I tried a little experiment today. I operate the CW Ops mini tests and normally use SKIMVALID in my filters that I use for cw skimmers....in this mornings test, I only saw one busted call K4TZ/TK4TZ. For this afternoons session, I changed SKIMVALID to NOT SKIMBUSTED and in about the same length of time (25-30 minutes), I ran into several more busts:
N7US/N7UN K4TZ/TK4TZ K5TU/TK5TU N2IC/TN2IC F5TKR/N5KTR N4AF/TN4AF
now this morning I was on 40M, and 20M this afternoon, so maybe prop played a role in it, and I know its a small sample set, but I am not convinced the NOT SKIMBUSTED works as well , or better, than SKIMVALID
------ Original Message ------
Date 2/3/2025 6:22:24 PM
Subject Re: [DXLab] Reasonable call prefilter
I posted this on the N1MM Logger+ group just over a month ago:
?
See <http://www.k3lr.com/w9zrx/AR-Cluster%20Filter%20Commands.pdf>, near the top of page 5.
?
Looking at those definitions: Skimvalid looks for multiple spots with the same callsign. Skimbusted looks for similar (e.g. "off-by-one") callsigns that are already marked as valid. A busted spot could still pass as Skimvalid if enough skimmers bust the call the same way, but if the correct call has already been spotted there, adding Not Skimbusted to the filter may help prevent the busted spot from showing up.
?
73,
Rich VE3KI
?
On Mon, Feb 3, 2025 at 05:04 PM, Jamie WW3S wrote:
Whats the difference between, if any, between NOT SKIMBUSTED and SKIMVALID ? My skimmer filters use SKIMVALID and other than occasionally getting a T in front of some calls, they seem pretty reliable …..
?
On Feb 3, 2025, at 7:41?AM, John K3TN via groups.io <jpescatore@...> wrote:
For all AR clusters you can enter a filter command that takes advantage of spot quality filter: SET DX FILTER NOT SKIMBUSTED
?
That does a great job of eliminating bad RBN/skimmer spots, based on work by CT1BOH. Not perfect, but really big step forward.
?
73 John K3TN
?
?
|
Here's the RBN? definition of the difference between SKIMVALID and NOT SKIMBUSTED:
?
"Skimvalid looks back to see if there are two or more spots of the same call, on the same frequency (+/- 0.2 KHz) before marking the spot as "Valid". Obviously, that will help with single busts. Skimbusted works by looking back to see if there is a similar but different spot, already determined to be Valid and posted on the same frequency (+/- 0.1 KHz). A similar call is one that can be transformed into the new call by character insertion, deletion, or substitution."
?
In practice, I use NOT SKIMBUSTED because it is is faster - often I get to the spot before the station has even finished CQing. In the long run, I spend less wasted time in established pileups and the occasional busted call is just one wasted click.
?
73 John K3TN
|
You may also want to add unique =
3 to your filter.
This makes sure that the call is heard at least three nodes.
Sounds simplistic, but if only one skimmer copies WC2L, then it
will not be sent to your system.
It can be as simple as SET DX FILTER NOT SKIMBUSTED AND
UNIQUE=3
Will WC2L
On 06-Feb-25 6:34 AM, John K3TN via
groups.io wrote:
Here's the RBN? definition of the difference between
SKIMVALID and NOT SKIMBUSTED:
?
"Skimvalid
looks back to see if there are two or more spots of the same
call, on the same frequency (+/- 0.2 KHz) before marking the
spot as "Valid". Obviously, that will help with single busts.
Skimbusted works by looking back to see if there is a similar
but different spot, already determined to be Valid and posted
on the same frequency (+/- 0.1 KHz). A similar call is one
that can be transformed into the new call by character
insertion, deletion, or substitution."
?
In practice, I use NOT SKIMBUSTED because it is is faster -
often I get to the spot before the station has even finished
CQing. In the long run, I spend less wasted time in established
pileups and the occasional busted call is just one wasted click.
?
73 John K3TN
--
William Liporace WC2L
or
AR-Cluster Node telnet dxc.wc2l.com 7373 or 144.93 MHz
wc2l@...
|
The ITU has fairly clear regulations on valid amateur callsigns
<>
essentially limiting callsigns to Letter-number (e.g. W1, G5, F2,
etc.), Letter-letter-number (e.g., WA1, KH6, PY3, VE2, etc.),
Letter-number-number (e.g. A61, C53, E77, etc.) or Number-letter-number
(e.g., 2E1, 3B8, 4L1, 4X4, 5A1, etc.) followed by up to four characters
the last of which must be a letter *except* in the case of half series
assignments (3DA-3DM Eswatani, 3DN-3DZ Fiji, SSA-SSM Egypt, and SSN-SSZ
Sudan) where the prefix should be *three* characters followed by a
number then a maximum of *three* characters the last of which must be
a letter.
*However* Article 19, section 3 includes language that permits the
use of callsigns with more than four characters in the suffix "On
special occasions, for temporary use".
DXKeeper permits logging callsigns of up to 13 characters (to handle
up to 7 character callsigns with a "portable" prefix or suffix). I
could not find a reference to maximum callsign size for SpotCollector.
For those who chase Special Event operations, what is a reasonable
(practical) limit for the length of special event (special occasion,
temporary use) callsigns that one might encounter on the air or see
via the cluster network? Is the 13 character limit in DXKeeper
workable (assuming there is no need for a "portable" modifier to
an already jumbo sized callsign <G>)?
73,
... Joe, W4TV
toggle quoted message
Show quoted text
On 2025-02-03 2:26 PM, Dave AA6YQ wrote: There are two reasons that I quickly implemented Bj?rn's suggestion:
1. it works ignores invalid callsigns from all spot sources, no matter what flavor of cluster management software they are running
2. if you're using the direct interoperation between DXLab and WSJT-X, it works ignores invalid callsigns decoded by local instances
of WSJT-X
73,
Dave, AA6YQ
|
Bj?rn SM7IUN
John K3TN
William WC2L
Jim N7US