[UK-CONTEST] Contest logger comparisons

[Jim Fisher]

The serial number problem has been bouncing around for ages.

With Win Test the serial can be stolen by the mult station before the Q is
logged by the run Op.

With other software using the serial number server it can't be stolen BUT

Quoting Larry F6FVY
"
We assume one RUN station and only one IN-BAND station.
We assume a perfect network.

T ! Stn RUN                        ! Stn IN-BAND
--!-----------------------------------------------------------------
1 ! Serial displayed : 001         ! Serial displayed : 001
2 ! CQ                             ! S & P
3 ! F6ABC replies to CQ            !
4 ! I enter F6ABC <Space>          !
5 ! So 001 is reserved             ! Serial displayed : 002
6 ! I enter the QSO                !
7 ! Serial displayed : 002         !
8 !                                ! I find and enter ZD8Z <space>
9 ! Serial displayed : 003         ! 002 is reserved
10! DL6ABC replies to CQ           ! I call ZD8Z but not heard
11! I enter DL6ABC                 ! Ditto
12! Serial displayed : 004         ! Ditto
13! JA6ABC replies to CQ           ! Ditto
14! I enter JA6ABC                 ! Ditto
15! Serial displayed : 005         ! Yeah ! I (finally) work him !
16! G6AA replies to CQ             !
17! I enter G6AA                   !
18! Serial displayed : 006         ! I find and enter CU2A <space>
19!                                ! Serial reserved : 006
20! Serial displayed : 007         ! I call CU2A (no reply)
21! ON6AA replies to CQ            ! Ditto
22! I enter ON6AA                  ! Ditto
23! Serial displayed : 008         ! Ditto
24! EA4AA replies to CQ            ! Finally CU2A is now QRT :-(
25! I enter EA4AA                  ! (F11)
26! Serial displayed : 009         ! Serial displayed : 009

Bottom line :

My final time-ordered log looks like :

F6ABC     001
DL6ABC    003
JA6ABC    004
ZD8Z      002
G6AA      005
ON6AA     007
EA4AA     008

1/ Serial numbers are not sequential (see ZD8Z)
2/ Serials are missing : No 006...
"

So neither solution is perfect.
There is a fundamental problem of sharing serial numbers across 2 machines
that can't be fixed by software.

73

Jim
GM0NAI / GM7R

-----Original Message-----
From: uk-contest-bounces at contesting.com
[mailto:uk-contest-bounces at contesting.com] On Behalf Of Steve Knowles
Sent: 18 June 2011 21:55
To: uk-contest at contesting.com
Subject: Re: [UK-CONTEST] Contest logger comparisons

If it results in an error in the sending log, which is in contravention of 
contest rules world-wide and which could result in unjust loss of points for

an innocent receiver, it is either a bug (which needs to be fixed) or a very

irresponsible piece of coding (which still needs to be fixed) ... view it 
how you will!

The RSGBCC has long been aware of the problem and of its likely link to 
certain logging software, rather than to allegations that the German Robot 
was altering entrants' logs in NFD and some of the other far-fetched 
suggestions which have been put forward from time to time.  But a suitably 
qualified statement of responsibility has been lacking .. until now!

RSGBCC policy is that, rather than make any automatic assumption of fault, 
adjudicators examine EVERY failed match in entrants' logs to ensure that 
fault is correctly ascribed.  For example, (and particluarly so on CW) it 
has been held for a long time that an error of one across a decade boundary 
(eg, 009 for 010) is unlikely to be due to the recipient ... the same across

a century boundary (eg 099 for 100) is extremely unlikely indeed to be a 
receive error, so the adjudicator may consider that the sending station is 
in error in such cases.  Now, adjudication software provides an instant 
analysis of ALL the mismatches involved with a particular station's log, 
enabling the adjudicator to make an informed judgement as to the quality of 
the sending operator at any time, spot any patterns of serial number or 
callsign error and identify sending errors over a wider number range.

If the adjudicator is satisfied that an error was made either by the 
transmitter or the receiver, then the side responsible for the error will 
lose the points.  If the adjudicator can not be sure who was responsible, 
then neither party loses any points.  Suggestions made by some competitors 
that they have lost large numbers of points unjustly are not valid, simply 
because each loss is individually considered.  That's why it takes so long 
to adjudicate the affected events!

Obviously, the recent information regarding the programs concerned with the 
logging errors may influence individual adjudicators' opinions but the basic

principles have not changed, nor is there any reason why they should.

73

Steve
G3UFY
Secretary RSGBCC



----- Original Message ----- 
From: "Oliver Bock" <dj9ao at gmx.net>
To: <uk-contest at contesting.com>
Sent: Saturday, June 18, 2011 3:06 PM
Subject: Re: [UK-CONTEST] Contest logger comparisons


> Hi Alan,
>
> i believe this refers to certain scenarios in a M/S environment. Wintest, 
> as N4ZR also notes,
> does not "reserve" a serial number for any of the stations running in such

> an environment.
> Due to that, I think, it may be that e.g. the running Operator sends a 
> particular number and
> before he enters the QSO a multiplier station on another band enters a QSO

> with exactly that
> number.
> The Win-Test developers have repeatedly explained that this is not a bug, 
> but a result of the
> particular way Win-Test handles M/S in its network.
>
> 73
> Oliver  DJ9AO/M0TAO
> _______________________________________________
> UK-Contest mailing list
> UK-Contest at contesting.com
> http://lists.contesting.com/mailman/listinfo/uk-contest 

_______________________________________________
UK-Contest mailing list
UK-Contest at contesting.com
http://lists.contesting.com/mailman/listinfo/uk-contest