Date   
Re: Results of Cluster Rating experiment

Michael Walker
 

Thanks Larry

Glad you found time to do this.

Mike va3mw


On Mon, Feb 17, 2020 at 3:47 PM Larry K8UT <K8UT@...> wrote:

On February 4th the N1MM Dev Team released an update containing an experimental Packet Cluster Rating feature. This feature ran for 7 days and reported 19,860 attempts by N1MM+ to connect to 675 Packet Cluster URLs. The Dev Team’s goal in this experiment was to determine whether desktop data collection could produce a valid cluster availability list.

FILTERS: Those 675 URLs were filtered to retain only 129 favorable sites.  Removed from the list were sites: with less than 10 successful connections; less than 75% successful connection rate; to localhost, 127.0 or 192.128 sites (internal LAN connections); to the reversebeacon backbone (which individual stations are not supposed to use).

GEOGRAPHY: To our surprise, location was a strong factor in identifying favorable sites. Sites that were popular on one continent were not used on other continents. Sites with high availability rates on one continent suffered from poor availability on other continents. For example, VE7CC.NET had a 94% success rate in NA, but only 76% from EU. DX.HU:9000 had 89% (EU), 77% (AS) and 100% (OC).

POPULARITY: Record totals by continents were: EU (9621), NA (7377), AS (1560), SA (618), OC (335), AF (120) and {surprise!} AN (0). A longer collection interval, or collecting during a major contest, might have produced better data from AF and OC. VE7CC.NET was the clear favorite with 25% of the filtered results – although all of its traffic originated in NA and EU. There were no filtered VE7CC records from the other continents. Unfortunately, the second most popular URL was the REVERSEBEACON.NET backbone, with 932 records – supporting the claim that these client connections to the backbone cause serious performance issues on the RBN network during major contests.

AVAILABILITY: The filtered list of 129 includes 46 that were available 100% of the time, and the average availability for all URLs was 93%. As already described, availability varied by continent. DXFUN.COM, for example, reported 96% (NA), 86% (EU) and 77% (SA). These variations are clearly not the fault of the cluster server, but due to the Internet infrastructure between the Ham’s QTH and the server site.

DOWNLOAD: If you would like to replace the default Telnet list in N1MM+ with this experimental list, you can download it from the N1MM+ Website’s >Downloads >Experimental Program Files folder. Copy the file to your local /users/…/Documents/N1MM Logger+/Export folder. In N1MM+, open the >Window >Telnet >Clusters tab. Click the [Edit List] button, and in the Edit Telnet List dialog window, click >File >Import. NOTE: The new file will overwrite your current cluster selection - write it somewhere before you import!

 

Finally - keep in mind that this is experimental. I’m still not sure how we’ll incorporate a verified cluster list into N1MM+.


-larry (K8UT)

Re: Connecting to Cluster

Michael Walker
 

Fred

As a test, try mine, va3mw.dxcluster.org port 41112.

Mike va3mw


On Mon, Feb 17, 2020 at 4:11 PM Tom Wagner (N1MM) <n1mmtomwagner@...> wrote:
1. NEVER connect to RBN.

2. Read K8UT's email from 20 min ago.

Tom - N1MM

On 2/17/2020 3:58 PM, Fred wrote:
> I am having trouble connecting to any cluster except for RBN.
>
> Any ideas on why this is happening?
>
> Thanks
> Fred KB4QZH
>



Re: Connecting to Cluster

Tom Wagner (N1MM)
 

1. NEVER connect to RBN.

2. Read K8UT's email from 20 min ago.

Tom - N1MM

On 2/17/2020 3:58 PM, Fred wrote:
I am having trouble connecting to any cluster except for RBN.

Any ideas on why this is happening?

Thanks
Fred KB4QZH

Connecting to Cluster

Fred
 

I am having trouble connecting to any cluster except for RBN.

Any ideas on why this is happening?

Thanks
Fred KB4QZH

Results of Cluster Rating experiment

Larry K8UT
 

On February 4th the N1MM Dev Team released an update containing an experimental Packet Cluster Rating feature. This feature ran for 7 days and reported 19,860 attempts by N1MM+ to connect to 675 Packet Cluster URLs. The Dev Team’s goal in this experiment was to determine whether desktop data collection could produce a valid cluster availability list.

FILTERS: Those 675 URLs were filtered to retain only 129 favorable sites.  Removed from the list were sites: with less than 10 successful connections; less than 75% successful connection rate; to localhost, 127.0 or 192.128 sites (internal LAN connections); to the reversebeacon backbone (which individual stations are not supposed to use).

GEOGRAPHY: To our surprise, location was a strong factor in identifying favorable sites. Sites that were popular on one continent were not used on other continents. Sites with high availability rates on one continent suffered from poor availability on other continents. For example, VE7CC.NET had a 94% success rate in NA, but only 76% from EU. DX.HU:9000 had 89% (EU), 77% (AS) and 100% (OC).

POPULARITY: Record totals by continents were: EU (9621), NA (7377), AS (1560), SA (618), OC (335), AF (120) and {surprise!} AN (0). A longer collection interval, or collecting during a major contest, might have produced better data from AF and OC. VE7CC.NET was the clear favorite with 25% of the filtered results – although all of its traffic originated in NA and EU. There were no filtered VE7CC records from the other continents. Unfortunately, the second most popular URL was the REVERSEBEACON.NET backbone, with 932 records – supporting the claim that these client connections to the backbone cause serious performance issues on the RBN network during major contests.

AVAILABILITY: The filtered list of 129 includes 46 that were available 100% of the time, and the average availability for all URLs was 93%. As already described, availability varied by continent. DXFUN.COM, for example, reported 96% (NA), 86% (EU) and 77% (SA). These variations are clearly not the fault of the cluster server, but due to the Internet infrastructure between the Ham’s QTH and the server site.

DOWNLOAD: If you would like to replace the default Telnet list in N1MM+ with this experimental list, you can download it from the N1MM+ Website’s >Downloads >Experimental Program Files folder. Copy the file to your local /users/…/Documents/N1MM Logger+/Export folder. In N1MM+, open the >Window >Telnet >Clusters tab. Click the [Edit List] button, and in the Edit Telnet List dialog window, click >File >Import. NOTE: The new file will overwrite your current cluster selection - write it somewhere before you import!

 

Finally - keep in mind that this is experimental. I’m still not sure how we’ll incorporate a verified cluster list into N1MM+.


-larry (K8UT)

Re: Corrupted Database, update

K9MA
 

Larry,

I do have many backups of that database, so only the last contest was affected.

To review, while trying to back up the database to a thumb drive during the contest, I did it backwards, copying the previous backup file into the database folder while the logger was running. I got an error message, but apparently that corrupted the database. However, the logger seemed to be working normally for the rest of the contest. When I tried to quit the logger (foolishly before exporting any files), it crashed and froze. I had to kill it with the task manager. A crash occurred every time I tried to open that log, even on another computer. I found that I could open other logs in the database successfully.  In retrospect, I should have exported an adif and the cabrillo file before quitting the program.

For others who might someday pull off a similarly boneheaded maneuver, here's what I did:

1. Renamed the corrupted database, so the program wouldn't try to start up with the corrupted log and crash.

2. Restored the contest log to a new database from the transaction log.

3. Exported that log to an adif file.

4. Deleted the affected log in the corrupted database.

5. Import the adif file to the (formerly) corrupted database.

The only hitch is that when I imported from the adif file into a new log with the same start date and time, the program insisted that there were already 1,631 contacts in it, though none showed up in the log window. After "allowing" dupes from the adif file, everything seemed to return to normal. The transaction log saved my bacon. I didn't have that backed up, but at least I didn't corrupt it, too.

Sorry for the bother, but thanks for the help.

73,
Scott K9MA


On 2/17/2020 13:44, Larry K8UT wrote:
Scott,

The SQLite Expert Pro "Repair" routine did not fix this, and it took a little more effort than usual. Your repaired database has been shipped back to you in a direct email.

As you suspected, the DXLog table (the table in the database holding the contacts) was corrupted by whatever-it-was that you did to the live file. I was able to restore all of your other logs and 1471 Qs from the ARRL CW contest. (not sure how many beyond 1471 were lost)

Your database contains 56289 contacts from the past two years. If you had completely trashed the database, you might have been very sad. I hope that you are using an appropriate backup program on your computer.

-larry (K8UT)

------ Original Message ------
From: "K9MA" <K9ma@...>
Sent: 2020-02-17 10:53:14
Subject: Re: [N1MM+] Corrupted Database, update

On 2/17/2020 05:26, Larry K8UT wrote:
Scott,

You used the right method for recovering a contest log: create a new database; >FIle >New Log the same contest type; and restore that contest into that new database from the transaction .TRN files. Well done.

RE: Your old database with one broken log - The Dev Team uses a tool called SQLite Expert Pro to repair databases. There is a free version... but it does not include the Repair feature. If you'd like, I'll attempt the repair on your broken file - send me any files in your /databases folder with that  filename and the suffixes .S3DB, .,S3DB-WAL, or .S3DB-SHM. (the s3db file will be there. the wal and shm temporary files may not exist... don't worry if they are missing)

-larry (K8UT)

------ Original Message ------
From: "K9MA" <K9ma@...>
Sent: 2020-02-17 03:40:19
Subject: Re: [N1MM+] Corrupted Database, update

Further update:

Apparently, only the log for the last contest was corrupted in the database. I was able to recover it from the transaction file to a new database. The old database is OK unless I try to open up that log. It was about time to start a new database, anyway. I'm not sure whether I can delete that log, but it doesn't really matter.

73,
Scott K9MA




On 2/17/2020 01:05, K9MA wrote:

I've just discovered that the database seems to be OK, UNLESS I try to open the contest which was open when it got corrupted. So the question reduces to how can I recover that one contest?

73,

Scott K9MA

-- 
Scott  K9MA

k9ma@...




-- 
Scott  K9MA

k9ma@...

Re: Mults problem in ARRL CW DX

Ron Stone
 

Indeed it is the PE / PEI issue, as I revised my log and submitted it. Now the correct number of mults is shown.

73
Ron GW3YDX

Re: Mults problem in ARRL CW DX

Steve London
 

The ARRL contest robot may not agree, but the correct abbreviation, as specified in the contest rules, is PE.

I have brought this issue to the attention of the person who supports the ARRL contest robot.

73,
Steve, N2IC

Re: Corrupted Database, update

Larry K8UT
 

Scott,

The SQLite Expert Pro "Repair" routine did not fix this, and it took a little more effort than usual. Your repaired database has been shipped back to you in a direct email.

As you suspected, the DXLog table (the table in the database holding the contacts) was corrupted by whatever-it-was that you did to the live file. I was able to restore all of your other logs and 1471 Qs from the ARRL CW contest. (not sure how many beyond 1471 were lost)

Your database contains 56289 contacts from the past two years. If you had completely trashed the database, you might have been very sad. I hope that you are using an appropriate backup program on your computer.

-larry (K8UT)

------ Original Message ------
From: "K9MA" <K9ma@...>
Sent: 2020-02-17 10:53:14
Subject: Re: [N1MM+] Corrupted Database, update

On 2/17/2020 05:26, Larry K8UT wrote:
Scott,

You used the right method for recovering a contest log: create a new database; >FIle >New Log the same contest type; and restore that contest into that new database from the transaction .TRN files. Well done.

RE: Your old database with one broken log - The Dev Team uses a tool called SQLite Expert Pro to repair databases. There is a free version... but it does not include the Repair feature. If you'd like, I'll attempt the repair on your broken file - send me any files in your /databases folder with that  filename and the suffixes .S3DB, .,S3DB-WAL, or .S3DB-SHM. (the s3db file will be there. the wal and shm temporary files may not exist... don't worry if they are missing)

-larry (K8UT)

------ Original Message ------
From: "K9MA" <K9ma@...>
Sent: 2020-02-17 03:40:19
Subject: Re: [N1MM+] Corrupted Database, update

Further update:

Apparently, only the log for the last contest was corrupted in the database. I was able to recover it from the transaction file to a new database. The old database is OK unless I try to open up that log. It was about time to start a new database, anyway. I'm not sure whether I can delete that log, but it doesn't really matter.

73,
Scott K9MA




On 2/17/2020 01:05, K9MA wrote:

I've just discovered that the database seems to be OK, UNLESS I try to open the contest which was open when it got corrupted. So the question reduces to how can I recover that one contest?

73,

Scott K9MA

-- 
Scott  K9MA

k9ma@...


-- 
Scott  K9MA

k9ma@...


-- 
Scott  K9MA

k9ma@...

Re: Flex3000+PowerSDR and SO2V operation - invalid self switching Monitore function

Tomáš Káčerek
 

Hi John, that is not about the MultiRx switching. I use PowerSDR with No RX2 radio Flex3000 few month and lot of contests i worked.

The problem is about MON (monitoring) own NF signal. It is needful to check and listen own CW keying.
Then MON button has three of stages. MON = monitore off, MONs = monitore on secondary stage processing, MONpr = minitore ON primary stage NF preprocessing. It is cycled by this three stages.

When I going to contact on VFO/B by "\" or mouse click into second entry window and when I going back to VFO/A then it's happened the situation, that MON function on transceiver going to status MONpr and I lost monitore sound from transceiver. After this I have to doubleclick on MON button in PowerSDR to reactive MONs status to hear a monitore CW own produced signal.

It is hardly complicate contest use of SO2V functionality in stereo mode VFOa/VFOb which I like to use on mz previouse transceires with dual VFO capability K3, FT1000 or with SDR 2nd receiver.

CAT commands which manipulates with this function are ZZM0, ZZM1 and ZZM2

I hope that it is clear. If not, I can send a video of use. It is simply reproducable issue.

Tom ok1dwq

Re: IC-9700 on 70 cm

Drew Vonada-Smith K3PA
 

Hi John,

Not sure what a transverter entry is (never used N1MM above 6M), but definately never created one.

73,

Drew K3PA

Re: Mults problem in ARRL CW DX

Ron Stone
 

Greetings guys

My good friend Dennis F5VHY may have put the finger on the problem. He did a SOSB 80 and believes that our contacts with VY2 have not qualified for a mult since ARRL expects PEI and both WinTest which he uses and N1MM+ expect PE. We all logged the VY2's as PE of course. Dennis points out that altering the log after submission is a disqualification offense so lets hope someone tells them about this !!

73
Ron GW3YDX

On Monday, 17 February 2020, 16:43:14 GMT, Steve London <n2ic@...> wrote:


The Cabrillo file sent to the contest robot contains no multiplier information.
It's just a list of your contacts and the info you entered. I have no idea what
an "N1MM Cabrillo app" is. No such thing. N1MM computes the multipliers using
the downloaded wl_cty.dat file, which is maintained outside of the N1MM team, by
AD1C.

It is always a good idea to do 2 things, just after the contest:

- Download the latest wl_cty.dat using Tools->Download and install latest
country file
- Rescore the contest using Tools->Rescore Current Contest

73,
Steve, N2IC

On 02/17/2020 09:29 AM, Richard King wrote:
> Well, this confuses me a little.
>
> Before I went to the ARRL contest robot, I created the Cabtillo file with the
> N1MM app. In that Cabrillo file it said my score was decreased by 6 multipliers.
> (Yes, I worked TO4A on 6 bands). But the QSO points was correct.
>
> Then I looked back at the N1MM summary window and verified that it said I had 6
> more mults and the same QSO points as the Cabrillo file it had just generated
> using N1MM. A look at the log file and it said TO4A was FM.
>
> So it would appear that the score reduction occurred before the Cabrillo was
> sent to the ARRL robot. Does the N1MM Cabrillo app somehow reference the ARRL
> web site BEFORE creating the Cabnrillo?
>
> 73, Richard - K5NA
>
> On Mon, Feb 17, 2020 at 4:08 PM Steve London <n2ic@...
> <mailto:n2ic@...>> wrote:
>
>    The problem is with the ARRL contest robot. The country file on the robot has
>    not been updated to include TO4A.
>
>    Relax....the log checker, a human being, will get it right in the final results.
>
>    73,
>    Steve, N2IC
>
>    On 02/17/2020 08:53 AM, Anthony W. DePrato wrote:
>      > Ken
>      > I got the same message they cut my score by over a thousand points.. have no
>      > ideal why was using N1MM
>      > 73 Tony WA4JQS
>      >
>      >
>      >
>      >
>
>
>
>



Re: Mults problem in ARRL CW DX

Steve London
 

The Cabrillo file sent to the contest robot contains no multiplier information. It's just a list of your contacts and the info you entered. I have no idea what an "N1MM Cabrillo app" is. No such thing. N1MM computes the multipliers using the downloaded wl_cty.dat file, which is maintained outside of the N1MM team, by AD1C.

It is always a good idea to do 2 things, just after the contest:

- Download the latest wl_cty.dat using Tools->Download and install latest country file
- Rescore the contest using Tools->Rescore Current Contest

73,
Steve, N2IC

On 02/17/2020 09:29 AM, Richard King wrote:
Well, this confuses me a little.
Before I went to the ARRL contest robot, I created the Cabtillo file with the N1MM app. In that Cabrillo file it said my score was decreased by 6 multipliers. (Yes, I worked TO4A on 6 bands). But the QSO points was correct.
Then I looked back at the N1MM summary window and verified that it said I had 6 more mults and the same QSO points as the Cabrillo file it had just generated using N1MM. A look at the log file and it said TO4A was FM.
So it would appear that the score reduction occurred before the Cabrillo was sent to the ARRL robot. Does the N1MM Cabrillo app somehow reference the ARRL web site BEFORE creating the Cabnrillo?
73, Richard - K5NA
On Mon, Feb 17, 2020 at 4:08 PM Steve London <n2ic@... <mailto:n2ic@...>> wrote:
The problem is with the ARRL contest robot. The country file on the robot has
not been updated to include TO4A.
Relax....the log checker, a human being, will get it right in the final results.
73,
Steve, N2IC
On 02/17/2020 08:53 AM, Anthony W. DePrato wrote:
> Ken
> I got the same message they cut my score by over a thousand points.. have no
> ideal why was using N1MM
> 73 Tony WA4JQS
>
>
>
>

Re: Mults problem in ARRL CW DX

Richard King
 

Well, this confuses me a little.

Before I went to the ARRL contest robot, I created the Cabtillo file with the N1MM app. In that Cabrillo file it said my score was decreased by 6 multipliers. (Yes, I worked TO4A on 6 bands). But the QSO points was correct.

Then I looked back at the N1MM summary window and verified that it said I had 6 more mults and the same QSO points as the Cabrillo file it had just generated using N1MM. A look at the log file and it said TO4A was FM.

So it would appear that the score reduction occurred before the Cabrillo was sent to the ARRL robot. Does the N1MM Cabrillo app somehow reference the ARRL web site BEFORE creating the Cabnrillo?

73, Richard - K5NA


On Mon, Feb 17, 2020 at 4:08 PM Steve London <n2ic@...> wrote:
The problem is with the ARRL contest robot. The country file on the robot has
not been updated to include TO4A.

Relax....the log checker, a human being, will get it right in the final results.

73,
Steve, N2IC

On 02/17/2020 08:53 AM, Anthony W. DePrato wrote:
> Ken
> I got the same message they cut my score by over a thousand points.. have no
> ideal why was using N1MM
> 73 Tony WA4JQS
>
>
>
>



Re: Mults problem in ARRL CW DX

Ron Stone
 

Well, well. well. 

Exactly the same here for my SOSB 160 entry. 

N1MM+ says 51 mults . Yet ARRL says only 50 mults. And boy have I checked them and checked them again to the point of right-clicking on the little blue tile in the mults table and establishing that the stations that were the mult are all present and correct and correctly in the log I have submitted. 

73
Ron GW3YDX - GW5R in the contest. 

ps - Thanks to all members of the group who gave me points for my GW5R SOSB 160 entry and also worked me as GW3YDX on other bands during the event. 

Re: Mults problem in ARRL CW DX

Steve London
 

The problem is with the ARRL contest robot. The country file on the robot has not been updated to include TO4A.

Relax....the log checker, a human being, will get it right in the final results.

73,
Steve, N2IC

On 02/17/2020 08:53 AM, Anthony W. DePrato wrote:
Ken
I got the same message they cut my score by over a thousand points.. have no ideal why was using N1MM
73 Tony WA4JQS

Re: using cut numbers in "Sent Exchange" when setting up a contest

CO8NMN Noe
 

Same to me but was starting the contest and my fastest answer was change the exchange value to ATT on the contest setup page in order to don't change the macros config. Hope the answer too best rgds. Noe CO8NMN


El lun, 17 de febrero de 2020 06:11, DD5XX <dd5xx@...> escribió:
Hi all,

when creating a new log for a contest there is a field "Sent Exchange". For example, in the ARRL DX CW contest I use "100". My macro is using the {EXCH} for sending. Unfortuantely this is not what I want to. Under [Config] --> [Configure Ports, Mode Control, ... Other] in the second tab [Function Keys] I have checked "Send cut numbers" and configured "Cut Number Style" = "(TAN)".

Unfortunately this doesn't seem to work because it shows no effect in a contest. Whenever I sent the exchange N1MM+ transmits the number 100, but I would have expected to hear a ATT (=short cut for 100).

I did use following workaround --> "File" --> "Open Log in database" --> tab [Contest] --> "Sent Exchange" = ATT
and after the contest I changed this setting back to the value 100.

But why doesn't this work as expected, what did I miss? any help appreciated. Thank you in advance.
 
 

 

 

Re: Can't Upgrade

n3nz Noel
 

Thanks Matt: your information was very helpful. Uninstalled N1mm and
re-install, the program is runnig with latest upgrade 1.0.8152.
Have a good day.

73, Noel(OM) N3NZ

Re: make spectrum window vertical

John Vogel
 


I’ve just reconnected the ic7300 and I wonder how this worked out?


On Feb 14, 2020, at 6:39 PM, Jamie WW3S <ww3s@...> wrote:

No, and I tried to grab a corner and resize but no joy....


On Feb 14, 2020, at 6:32 PM, John Vogel <johnvogel74@...> wrote:

I’m back home - did you get it?  
I can connect up the 7300&laptop and compare notes?


On Feb 14, 2020, at 3:28 PM, Jamie WW3S <ww3s@...> wrote:


I see column and bar, but no matter what I chose, it remains horizontal which I assume is BAR. 

------ Original Message ------
From: "John Vogel" <johnvogel74@...>
Sent: 2/14/2020 3:12:20 PM
Subject: Re: [N1MM+] make spectrum window vertical

It’s an option selected in the upper right hand corner of the spectrum window.

John 
N1PGA 


On Feb 14, 2020, at 3:07 PM, Jamie WW3S <ww3s@...> wrote:

 I thought I saw a recent post about how to make the spectrum window vertical, I cannot locate it now.....does anyone have it? 

Re: Corrupted Database, update

K9MA
 

On 2/17/2020 05:26, Larry K8UT wrote:
Scott,

You used the right method for recovering a contest log: create a new database; >FIle >New Log the same contest type; and restore that contest into that new database from the transaction .TRN files. Well done.

RE: Your old database with one broken log - The Dev Team uses a tool called SQLite Expert Pro to repair databases. There is a free version... but it does not include the Repair feature. If you'd like, I'll attempt the repair on your broken file - send me any files in your /databases folder with that  filename and the suffixes .S3DB, .,S3DB-WAL, or .S3DB-SHM. (the s3db file will be there. the wal and shm temporary files may not exist... don't worry if they are missing)

-larry (K8UT)

------ Original Message ------
From: "K9MA" <K9ma@...>
Sent: 2020-02-17 03:40:19
Subject: Re: [N1MM+] Corrupted Database, update

Further update:

Apparently, only the log for the last contest was corrupted in the database. I was able to recover it from the transaction file to a new database. The old database is OK unless I try to open up that log. It was about time to start a new database, anyway. I'm not sure whether I can delete that log, but it doesn't really matter.

73,
Scott K9MA




On 2/17/2020 01:05, K9MA wrote:

I've just discovered that the database seems to be OK, UNLESS I try to open the contest which was open when it got corrupted. So the question reduces to how can I recover that one contest?

73,

Scott K9MA

-- 
Scott  K9MA

k9ma@...


-- 
Scott  K9MA

k9ma@...


-- 
Scott  K9MA

k9ma@...