WeatherFax ...

[HUR450502]

.... ok, here is the image file and the track (which are now .zip.doc files), so rename and unzip... ok??

Will try to get the other pieces split up enough to transmit them as well, ok??

Best regards, Uwe

[HUR450502]

.... and here the OpenCPN Directory (allusers/OpenCPN) - again, faked as a .doc file..... OK??? Here the plugin subdirectory is incuded as well.....

Best regards, Uwe

[CarCode]

Uwe,
obviously there is a bug in weatherfax which changes the scale factor in opencpn.ini so you get this large scale factor after starting OpenCPN again (step 7).
Gerhard

[rgleason]

Uwe, you have my attention.

[HUR450502]

Rick, one more point: I looked at the navobj. files and in the data you have, the navobj.xml.1 contains the data for the "correct" track - means, if you would rename it to navobj.xml, you have the correct track again. The current navobj.xml has the problem, so, by comparing them, there might be some indication, what has happened. What do you think? Regrettably I didn't save the log between the diferent attempts. However, if this would help, I could take a copy of this directory set between every single step..... OK?? - Just let me know....

Best regards, Uwe

[rgleason]

Uwe Good point. I think one or two examples would be good.

Sent from my SCH-I545 using Cruisers Sailing Forum mobile app

[rgleason]

Carcode that could be the result. Something to do with resolution.

Sent from my SCH-I545 using Cruisers Sailing Forum mobile app

[HUR450502]

...... o.k. Rick,

here we go: You receive 9 ".zip.kml" files (some were too big to use the .doc type....), which are the contents of

c.\users\allusers\opencpn

As I do not know, whether or not OpenCPN writes to the datasets during processing, I have saved some of the directories while OpenCPN was running.

So, here is the sequence (file number 1 - 9):

1. Right after installation and first start
2. After installation and enablement of the weatherfax plugin
3. After import of the track
4. OpenCPN running - before calibration
5. OpenCPN running - after calibration (the map displays fine with overlay and track)
6. Program closed
7. Program re-opened and running - displays "Overzoom"
8. Zoom - out to display the full (now distorted) track - still running
9. Program closed

What do you think.... I believe, this is as close as we can get without any debugging or tracing tool....

OK???

Have fun!!

Best regards,

Uwe

[rgleason]

Uwe, this looks good. I hope to get to this in 2 days, having technical problem with a website.

[HUR450502]

Hi Rick,

I looked a little bit around - mainly because of "carcode's" comments, and yes, after calibration there are parameters changed in the .ini file.

The "vpscale" looked funny, instead of 0.00015 - where I left it, it now changed to some (don't xactly remember the correct value, however) "exponential number" like 8,257 E-0035, or similar. I think the exact value is not interesting, but this extraordinary small scale-factor, as well - and that concerns me - that there is a decimal "comma" instead of a decimal point.

Further the vpLatLon changes such, that the latitude is exactly on a full degree (in my case 42.0000) and the longitude is set to exactly the "ownshipLatLon" longitude. Also the latitude of the "ownshipLatLon" is set to a full degree value - in my case 33.0000

So, these parameters look:

[Settings/GlobalState]
VPLatLon=" 42.0000, -79.2820"
VPScale=0.00015
VPRotation=0
OwnShipLatLon=" 33.0000, -79.2820"
bFollow=0

before they were on some "exact" values like 42.2354 and 33.3580 .

This is also, what I see on the track which is on "full" degrees only..

Does this help a bit, so you don't have to look up everything??

Best regards and have a nice wekend,

Uwe

[rgleason]

Uwe, it looks like you are figuring out the details of the symptoms quite readily. I think Sean would ask me to compile the program in debug mode and put it through the steps you outline, making the program step through somehow, until we saw where in the code that it does those transformations to your Track!

Then having identified the culprit area of the code, he could address the problem.
Just thinking about this right now. I will need to figure out how to set break points etc in debbug mode.

[HUR450502]

hi Rick and Sean,

yes...... of course. If I can be of help, I will "do my very best .... :-) "

One more observation, which might narrow this down: When I do the calibration and then import the track (the overlay is still on the map) the track is distorted as well. If I close the weatherfax plugin so the overlay disappears, and then import a track it is distorted as well. If I close the program and reopen again and then import the track, it will be fine again....

In the screenshot below you see the display with the track imported right after calibration and then the smooth lined track imported after restarting the program.....

Nice, isn't it???

If it helps??

Best regards,

Uwe

[rgleason]

Uwe reports:
1. Calibrate the image completed and overlayed on chart, then import the track, then track gets distorted.
2. Close weatherfax and overlay disappears, then import the original track and the track is distored again.
3. Close the program and reopen, then import the original track with no distortion.

Uwe, is this correct interpretation?

[HUR450502]

Hi Rick,

that is correct: This way I had three tracks the two (distortd) tracks, that overlaid on each other and then - after restart of OpenCPN (without any weatherfax startet etc. - the undistorted track. This is what the screenshot shows.

OK?

Best,

Uwe

[rgleason]

Uwe,
You have provided ample documentation.
the weatherfax TIFF file
the track file
your navobj files.
the sequence used to create the stepped modified track.
screenshots off the anomaly track that appears to have been rewritten
further documentation that certain scales have been rewritten
all having been done with a completely virgin install.
your investigation has been thorough and well documented.
I do not doubt that the modification/change of the track event occurs.

However, try as I might, I have not been able to reproduce it this morning. I am sorry Uwe, I fully expected to have a failure.

I have done the following:
Loaded the gpx you provided.
Loaded the tiff you provided, calibrated it and checked the track, no change.
Then shutdown, and reloaded. no change.

Loaded the tiff, calibrated.
Loaded the track, no changed.
Shutdown weatherfax.
Opened weatherfax. Tried edit recalibrate, no change.
Shutdown O. reopen.
Open weatherfax. load tiff, calibrate.
Load track. no changed.
Tried it a dozen different ways. No change in track.

I am wondering what you settings are for the calibration?
Please see mine in the attached screenshot. Are they terribly different?
I could upload a bunch of screen shots of the first steps process but they would not tell you anything different...
We need to get to the nub of this problem somehow....

It occurs to me that you provided the coordinates.xml
Perhaps that has some clue....
Will check that later..

Attached are in no particular order, just show some of the process.
Any suggestions? Standing by.