Weather Routing

[rgleason]

It appears that Boat_Polar.CSV files are saved out and opened successfully.

However, there are now some problems WeatherRoutingPositions
1. They do not show up in the Route Manager Waypoints anymore.
2.When they are deleted they remain in the dropdown under Configuration.
3. They do not show up on the chart in any form with any mark or designation after they are created..

[seglar-lulle]

I'm on holiday with family without both boat and computer so I can't do tests. In a couple of weeks I will look into this again.

Sent from my GT-I9505 using Cruisers Sailing Forum mobile app

[tom_BigSpeedy]

Hi Rick,

it's not just the beam, that is not saved.
I enter my boat settings, like in the screenshot, then hit the button "Save and Close" (boat_settings.gif)

Next I simply re-open the whole boat config with the Edit-button(reopen_boat_settings.gif), and all settings are gone again (boat_settings_after_reopen.gif), only hull-speed is still wrong ...

By the way : hull speed calculation (see screenshots) must be wrong :
a 29 ft waterline can't have 10.8 knots of hullspeed.
The basic, known formula shows ~7.3, which is correct in real life.
You can't fake reality, although I'd wish so sometimes...
Or it's not hullspeed in the known sense, but something else ...

By the way : is sail area meant with white sails only, or including Gennaker ?
I assume mainsail and solent only ...

Since a couple of weeks, the plugin rejects to load my polar csv files correctly. They still worked fine in June.
No, there are no spaces at then end of the lines, something else must have changed.
Did you revert rows/columns ?

Would be nice to know first hand what has changed instead of keeping beta testers busy with reverse engineering.

Will stop testing until this is resolved.

Thomas

[seandepagnier]

Originally Posted by tom_BigSpeedy

Hi Rick,

it's not just the beam, that is not saved.
I enter my boat settings, like in the screenshot, then hit the button "Save and Close" (boat_settings.gif)

Next I simply re-open the whole boat config with the Edit-button(reopen_boat_settings.gif), and all settings are gone again (boat_settings_after_reopen.gif), only hull-speed is still wrong ...

I fixed very recently.

By the way : hull speed calculation (see screenshots) must be wrong :
a 29 ft waterline can't have 10.8 knots of hullspeed.
The basic, known formula shows ~7.3, which is correct in real life.
You can't fake reality, although I'd wish so sometimes...
Or it's not hullspeed in the known sense, but something else ...

It's because it's using a more modern formula which takes displacement into account. It is well known that very light efficient craft can go faster than the standard calculation without planing.

Try making your boat weigh enough tons.

By the way : is sail area meant with white sails only, or including Gennaker ?
I assume mainsail and solent only ...

It would be all sails, but this is only half-baked.

Since a couple of weeks, the plugin rejects to load my polar csv files correctly. They still worked fine in June.
No, there are no spaces at then end of the lines, something else must have changed.
Did you revert rows/columns ?

Please attach a polar file that doesn't work. I changed the way it stores the data to be more efficient and accurate.

I was working toward supporting sea state polars, but I don't think I can finish this soon, because it's a hard problem.

In the meantime polar files are the best option, so I would like to fix that.

[tom_BigSpeedy]

Originally Posted by boat_alexandra

I fixed very recently.

I used latest github sources and compiled.

Originally Posted by boat_alexandra

It's because it's using a more modern formula which takes displacement into account. It is well known that very light efficient craft can go faster than the standard calculation without planing.

Try making your boat weigh enough tons.

It doesn't help, she only weighs 3 tons, maybe 3.1 or 3.2 ...
The max. I ever reached upwind (40° app., @12 kts of true wind), with full water ballast, no wave and totally overpowered was ~8knots. But that's it. She'll never ever do 10.8, even when planing with a 120sqm Gennaker that's not easy to reach.

Originally Posted by boat_alexandra

Please attach a polar file that doesn't work. I changed the way it stores the data to be more efficient and accurate.

I was working toward supporting sea state polars, but I don't think I can finish this soon, because it's a hard problem.

In the meantime polar files are the best option, so I would like to fix that.

Here you go, see attachment (remove.doc suffix from file)
Thomas

[rgleason]

Thomas,

Thank you for beta testing it is very very helpful, because I am not always right about these things and Sean needs at least several of us guinea pigs to tell him what is wrong.. I was going to go back again and test the plugin because I don't understand some of the things I found....Using the most recent beta, However I want to assure you that the plugin really does work and you can do routing with it if you first do a few simple things. I got screens just like yours at first.

1. First I found that I needed to enter the correct boat data and compute the polars, then save a polar.csv file out. (Also loading that file worked ok.) However older files caused Opencpn to halt with sending a report.

2. To get my older polar working, I looked at the one that was saved for format, and found that I had to add rows of zeros (save out a polar csv file to see the format) above the max tacking angle to get my polars to be accepted. I am sorry, I forgot to mention that.

3. I am having trouble understanding how the file system works for boat_configuration_data.xml See http://www.cruisersforum.com/forums/...ml#post1593482
The data seems to be saved to two files, boat.xml and yourname.xml but if for example you save out beam="10" it does not seem to make it back to the plugin when you load the file..... Also often Sean's Bristol Boat_Alexandra data comes back in. I need to understand this situation better.

4. I also found some persistent anomalies writing and reading files and I wanted to understand that better. This has been a problem for me for awhile and I would put that first on the list for fixes.

5. There are now some problems WeatherRoutingPositions
A. They do not show up in the Route Manager Waypoints anymore.
B. When they are deleted they remain in the dropdown under Configuration.
C. They do not show up on the chart in any form with any mark or designation after they are created.. until a routing is done using those marks. (Is this due to change in the main beta program?)

6. The Batch mode works very nicely and can be a big help deciding when to leave.

7. I was just getting into understanding how the different Sail Plans and Switch Plan Rules work. Sean says that he is still working on that part too.

5. I was going to uninstall weather_routing from all versions of Opencpn ( I have about 3 versions running) and then delete all weather_routing directories, and confirm all weather_routing/data files are deleted. Then I was going to install a new fresh version to the current beta and test.

6. Before starting developing the wave polars I'd really like to suggest that we get a version running in windows that has all of the current features working. Of course this is up to Sean.

7. I think it is important to try to keep the default install simple so that it works for people who don't understand how the plugin works...such that their experience the first time is positive and they get some results.

[tom_BigSpeedy]

Rick,

you're absolutely right. We need a stable version first. Wave polars are another topic.

I'd be absolutely happy (for live testing), if I could simply use my own polar and do the basics. But currently that fails, due to the reasons you already mentioned.

Actually I was happy to get at least ONE polar for my boat, covering the most frequent windspeeds. Everything else increases the effort to record and generate them tremendously. I'm not a professional sailor and I don't have the time to do so.
The polar I uploaded this morning is also not correct, as I used polar_pi to create it from VDR files. And if you run polar_pi with VDRs it takes only the max values, instead of averaging (as it does with logbook-konni files, or as polauto does). But this polar is still better and more accurate than the calculated ones ...

Also I think Sean pushes everything to github, sometimes even w/o testing at all. Of course he can do, it's his project.
But then we (beta-testers) should change our habits and not pull every new commit from git.
That's what I'm currently doing in the good old O-manner, where I generally can rely on, that the code is throughly tested by the designers and then reviewed again by Dave, before it's deployed to the official O master.
And that's where I got trapped several times now with our plugin :

• a couple of times the makefiles didn't compile anymore,
• variables were not initialized and caused compiler errors,
• Sources were built for wxWidgets 3.0 (but didn't compile on the official 2.8.x)
• Source code witten that caused VC2010 compile errors
• then it compiles w/o errors, but crashes immediately when doing some basics, like trying to open a new configuration window, or opening your own polar file, which worked fine in the past ...

Please don't missunderstand : I don't want to blame Sean here at all.
He's doing a great job, and I'd love to see (t)his plugin working.

Current situation for me is since a couple of weeks, that I can't (sail and) test it at all, as it does compile (so I overwrote my existing, working dlls) but crashes/fails every now and then. And at the moment I simply don't have the time to dig deep into it, and find out -- step by step -- what's wrong, as you do (Good job, much appreciated).
Writing this comment is only possible, as I have a day off and it's raining outside

Maybe you should coordinate kind'a beta-releases when it's worth downloading/compiling, and announce them via the forum ...?

Cheers,

Thomas

[rgleason]

@Thomas
The plugin does work, just do the first few suggestions I describe above to get polars right! In fact we now have all the plugins compiling in MSVC++ (celestial nav, weatherfax, weather_routing, climatology, etc) which is pretty darn good) IMHO

Sean,
I am going to try to do the cleanup of weather_routing remnants on my various Opencpn programs and reinstall and then do a step by step review and summarize that in the next few days.

I think you probably know Dave Cowell
https://github.com/cowelld
https://github.com/canboat/BR24radar_pi

He seems to have an interest in boat performance too, maybe it would help to join efforts on some of this, discuss or start working towards a more data input nmea oriented Polar? -It might help take some of the load off you too, juggling multiple plugins plus optimization (which you've gotten integrated into the main program) --all big achievements.

https://github.com/canboat/BR24radar...mment-50474845

... I've been working on a new plug-in that is involved with analyzing boat performance. In SeaClear there's an add-on called SeaTrace and SeaTrackEdit which were written by another Dutch sailor with whom I've been corresponding.
Best, Rick

[tom_BigSpeedy]

Originally Posted by boat_alexandra

Please attach a polar file that doesn't work. I changed the way it stores the data to be more efficient and accurate.

...

In the meantime polar files are the best option, so I would like to fix that.

Sean,

I found an old version of the plugin, which I used during a race in June.
Please see the attached screenshot. That's what the polar file I uploaded yesterday looked like in May/June.

Thomas

[seandepagnier]

Originally Posted by rgleason

Thomas,

1. First I found that I needed to enter the correct boat data and compute the polars, then save a polar.csv file out. (Also loading that file worked ok.) However older files caused Opencpn to halt with sending a report.

I will try to fix with the file uploaded.

2. To get my older polar working, I looked at the one that was saved for format, and found that I had to add rows of zeros (save out a polar csv file to see the format) above the max tacking angle to get my polars to be accepted. I am sorry, I forgot to mention that.

3. I am having trouble understanding how the file system works for boat_configuration_data.xml See http://www.cruisersforum.com/forums/...ml#post1593482
The data seems to be saved to two files, boat.xml and yourname.xml but if for example you save out beam="10" it does not seem to make it back to the plugin when you load the file..... Also often Sean's Bristol Boat_Alexandra data comes back in. I need to understand this situation better.

Can you break it down into steps to reproduce?

4. I also found some persistent anomalies writing and reading files and I wanted to understand that better. This has been a problem for me for awhile and I would put that first on the list for fixes.

5. There are now some problems WeatherRoutingPositions
A. They do not show up in the Route Manager Waypoints anymore.

and never did

B. When they are deleted they remain in the dropdown under Configuration.

really? this must be a bug, I will fix.

C. They do not show up on the chart in any form with any mark or designation after they are created.. until a routing is done using those marks. (Is this due to change in the main beta program?)

Are they supposed to? How do you know the difference between start and end?

7. I was just getting into understanding how the different Sail Plans and Switch Plan Rules work. Sean says that he is still working on that part too.

Still has some bugs I think, but working much better now.

I just added support for colors to see when climatology data is used instead of grib, and also the map fades out when the data is deficient.

6. Before starting developing the wave polars I'd really like to suggest that we get a version running in windows that has all of the current features working. Of course this is up to Sean.

ok

7. I think it is important to try to keep the default install simple so that it works for people who don't understand how the plugin works...such that their experience the first time is positive and they get some results.

How can we make it simpler?

Originally Posted by tom_BigSpeedy

Rick,

you're absolutely right. We need a stable version first. Wave polars are another topic.

Alright, we will move in that direction.

I'd be absolutely happy (for live testing), if I could simply use my own polar and do the basics. But currently that fails, due to the reasons you already mentioned.

Actually I was happy to get at least ONE polar for my boat, covering the most frequent windspeeds. Everything else increases the effort to record and generate them tremendously. I'm not a professional sailor and I don't have the time to do so.
The polar I uploaded this morning is also not correct, as I used polar_pi to create it from VDR files. And if you run polar_pi with VDRs it takes only the max values, instead of averaging (as it does with logbook-konni files, or as polauto does). But this polar is still better and more accurate than the calculated ones ...

Yes, the calculated ones are still way off, it is a very difficult problem.

Also I think Sean pushes everything to github, sometimes even w/o testing at all. Of course he can do, it's his project.

Many times I make a mistake, but I almost always do test the changes first.

But then we (beta-testers) should change our habits and not pull every new commit from git.

I'm sorry for the breakage. You can always checkout older commits until you don't have problems anymore (then please post which commit it was)

That's what I'm currently doing in the good old O-manner, where I generally can rely on, that the code is throughly tested by the designers and then reviewed again by Dave, before it's deployed to the official O master.

we are lucky to have dave to keep the project together.

And that's where I got trapped several times now with our plugin :
[*]a couple of times the makefiles didn't compile anymore,

Makefiles don't compile

[*]variables were not initialized and caused compiler errors,

This should not cause a compiler error, maybe a runtime error

[*]Sources were built for wxWidgets 3.0 (but didn't compile on the official 2.8.x)

Yes, well I am trying to build with both versions, and it is annoying. wx2.8 doesn't work on macosx for example. In the future we should move to wxWigets 3.1. For now I am using 2.8 and for the immediate release. Keep in mind, everyone would be much better with the latest version, and I really wish we would switch.

[*]Source code witten that caused VC2010 compile errors

This will continue as I do not use this compiler. I do however compile with mingw for windows which works fine. Visual studio doesn't support a lot of features gcc has, which has been annoying and forces me to write code which runs slower, more complicated and hard to maintain. It is not a free compiler while gcc is and we are developing free software. I don't know why anyone wants to use visual studio, it would be much easier to stop supporting all of its differences.

By comparison, I tried llvm which is a new emerging compiler not long ago, and it compiled opencpn without errors much faster than gcc and the program ran perfectly.

[*]then it compiles w/o errors, but crashes immediately when doing some basics, like trying to open a new configuration window, or opening your own polar file, which worked fine in the past ...

Sorry I will try to fix these bugs. A backtrace would be very useful.

[COLOR=Red]Please don't missunderstand : I don't want to blame Sean here at all.
He's doing a great job, and I'd love to see (t)his plugin working.

Most of the problems are my fault, not intentional. I try always for improvements and feedback is essential.

Current situation for me is since a couple of weeks, that I can't (sail and) test it at all, as it does compile (so I overwrote my existing, working dlls) but crashes/fails every now and then. And at the moment I simply don't have the time to dig deep into it, and find out -- step by step -- what's wrong, as you do (Good job, much appreciated).
Writing this comment is only possible, as I have a day off and it's raining outside

Maybe you should coordinate kind'a beta-releases when it's worth downloading/compiling, and announce them via the forum ...?

Cheers,

Thomas

My plan is to first fix all of your problems, and release it.


For polar calculation, this is a big problem. I think maybe a separate application or plugin can be developed. My idea is something like this:

Enter all of your boat parameters, like sail types, sizes, boat sizes, hull type etc. Now take measurements with the gps, maybe just a few, and enter wind speed, or record from nmea. The program can convert to true wind speed/direction from apparent wind measurements. Also enter water temperature because cold water has steeper waves. Air temperature also because cold air being denser has more force at a given speed.

Now, with just a few measurements, the program can calculate several parameters related to sail efficiency, wave drag, and frictional drag. This will "calibrate" the computation to give correct results. With more measurements it becomes more accurate, but trying to get every possible wind speed and angle is difficult.

Would it be useful to plot the polar in 3 dimensions and allow the user to rotate it to see all the data at once?

[tom_BigSpeedy]

Hi Sean,

I don't want to deepen this ...
CMAKEs had changed and as a result it didn't compile anymore.
Last thing I remember was a bzlib-directory problem ...
Sorry my fault, I should have written : variables were not declared (instead of initialized).
But let's get back to work :

With new Boat.xml and WeatherRoutingConfiguration.xml, although loading "my" polar file it looks like this : screenshot1.gif
This is known and you're working on it.

Now, change Plot Coordinates from "Polar" Display to "Rect" (screenshot2.gif Edit TR:, the one showing up at the very bottom), and immediately back to "Polar"
Then the polar display changes to screenshot3.gif

Go to tab "Sail Plan Polar", and reload my polar file.
Polar display changes again (Screenshot4.gif) ...


Starting with a new Boat.xml, entering my boat data in tab Boat configuration, saving the configuration (button "Save and Close"), closing everything, restarting O does not help. My boat config parameters are all gone again ...


Hope that helps.

Thomas

[rgleason]

Tom Big Speedy,

I will add a directory to Opencpn File Thingie rguser rgpass called "Earlier PI" and move the outdated install files into that, so anyone can backup easily.

I'll keep 3 or 4 previous versions max in there. Also I will try to find some of the earlier compiles of weather_routing install and upload them. (I recently cleaned up my Opencpn files, so I may come up a little short on that (:-).
We really need your help (and others help) beta testing and I really appreciate the effort, so anything I can do to help you, let me know. (Ah, by your recent post it looks like you are Linux, so the above won't help. But I'll do it anyway.)

Sean has recently been working straight out on an interesting project (check out his github page) and he is juggling many programming projects, plus he has had his computer and backup computer break a little while ago, so you can imagine why he uploads to git regularly. The sea environment is what it is...I am sure you know that. I think we would all prefer he preserves his work rather than lose it.

Maybe we should have some signal in the Git message header that says when he his intent is to have us compile? Perhaps just a "Compile." at the beginning? Then you won't over write your previous with a version that is not intended for compiling and Sean can still use it to backup his programs too. I think this would be a good thing to do with the other plugins too.

Also I have some description for walking git back if you would like me to post it. It is actually quite easy but intimidating the first time. Let me know if you need it.

--
Sean, Thank you.

I know how it is difficult to compile MSVC and am ready to move on to 3.0 or 3.1 widgets, and am appreciative of the efforts to get plugin running smoothly.

I can certainly download the Mingw compiler and use that if it would be better, but do you think the results can be successfully used by other Windows users running Opencpn compiled on MSVC, ...OK?

2. To get my older polar working, I looked at the one that was saved for format, and found that I had to add rows of zeros (save out a polar csv file to see the format) above the max tacking angle to get my polars to be accepted. I am sorry, I forgot to mention that.

3. I am having trouble understanding how the file system works for boat_configuration_data.xml See Weather Routing
The data seems to be saved to two files, boat.xml and yourname.xml but if for example you save out beam="10" it does not seem to make it back to the plugin when you load the file..... Also often Sean's Bristol Boat_Alexandra data comes back in. I need to understand this situation better.

Can you break it down into steps to reproduce?

Yes, will do in next day.

Best, Rick

[tom_BigSpeedy]

Originally Posted by rgleason

Tom Big Speedy,

Maybe we should have some signal in the Git message header that says when he his intent is to have us compile? Perhaps just a "Compile." at the beginning?

Rick,

"branching" is the magic out-of-the-box-functionality git offers for your attempt ...
Or in other words :
keep the compile versions on master, and create for each new development a separate branch.
There you can fiddle (and back up) as long as you want. When you're done, merge to master.
Should be fairly easy, as Sean is the only one committing to weather_routing master at the moment (there's no risk that his dev-branch runs into compatibility problems with master)

And we (testers) would only download master, unless someone explicitly tells us to temporarily try and help out on another branch ...

Good idea to back up old versions. Makes it easier to compare between versions, or verify if we've lost something ...


Thomas

[seandepagnier]

Originally Posted by rgleason

I know how it is difficult to compile MSVC and am ready to move on to 3.0 or 3.1 widgets, and am appreciative of the efforts to get plugin running smoothly.

I can certainly download the Mingw compiler and use that if it would be better, but do you think the results can be successfully used by other Windows users running Opencpn compiled on MSVC, ...OK?

Yes, will do in next day.

Best, Rick

Rick, those users have to use the mingw compiled version of opencpn. You can test it with all the plugins here:

opencpnplugins - Browse /opencpn_mingw_experimental at SourceForge.net

Originally Posted by tom_BigSpeedy

Rick,

"branching" is the magic out-of-the-box-functionality git offers for your attempt ...
Or in other words :

Yes, the problem is I think what I've done is an improvement most of the time.

Also the cases where I think I can finish something sooner than I can.

But yes, I use a branch for longer term changes, just not so much in the plugins yet, but now with the increased feedback I will try to if needed.

[rgleason]

If the plugin is closing Opencpn and trying to send a failure report:

1. Close opencpn
2. Go to C:\ProgramData\opencpn\plugins\weather_routing and delete or move to another directory the boat.xml and weatherroutingconfiguration.xml, any other xml file and any *.csv file.
3. This plugin requires Climatology data and/or Grib data so make sure those plugins are installed.
4. Start Opencpn, Climatology, Grib with a recent grib file, and then start Weather_routing_pi
5. Enter two reasonably distant weatherrouting points by locating cursor near Cuba then right mouse click, pick "weather route position" and name "1". Locate cursor off end of Africa, right mouse click pick "weather route position" and name "2".
6. At the Weatherrouting menu pick Configuration > New , two menus will pop up instantly "Weather Routing Configuration" and then "Boat" menu on top with the focus, with the polar diagram messed up.
7. Click on the "Sail Plan Polar" Tab and the at the dropdown change "Loaded from CSV file" to "Computed from Sailboat Transform" and the diagram should change.
8. On the VMG Tab, uncheck optimize tacking (I've never understood this.)
9. Then pick "Save and close" which at that moment creates Boat.xml file in C:\ProgramData\opencpn\plugins\weather_routing. (I watched the file appear. There is no "weatherroutingconfiguration.xml yet in the directory.)
10. Now the focus is on the "Weather Routing Configuration" menu.
    
11. Make sure Start is Position "1" and End is Position "2".
12. Set the time step to 24 hours. Make sure grib is checked, Climatology is "Most likely" and check "Detect Land"
13. Make sure Integrator is set to "Newton's Method" (faster) and that Constraints is 90, 180, 120, 100.
14. Under Courses relative to true, Generate is 0,180, 5. Then hit "Generate". A bigger step is less accurate but speeds up processing.
    
15. Now "Close" the "Weather Routing Configuration" menu.
    
16. Note that no weatherroutingconfiguration.xml appears in C:\ProgramData\opencpn\plugins\weather_routing although opening menu is labeled that!
    
17. Now highlight the route in the menu and pick "Configuration" > "Compute"
18. Isochrones and a routing should appear in blue with an instantaneous alternate route in yellow connected to the cursor.
19. Close weather routing by clicking on the icon (There is no close button at the top right on the menu).
20. Note that no weatherroutingconfiguration.xml appears in C:\ProgramData\opencpn\plugins\weather_routing
21. Close Opencpn
22. Now a new file weatherroutingconfiguration.xml appears in C:\ProgramData\opencpn\plugins\weather_routing

Initialization -I thought that if:

1. No viable files were found with user names for user_boat.xml and user_weatherconfiguration.xml,
2. Boat.xml and Weatherroutingconfiguration.xml did not exist

Then in that event I thought the plugin created them when it started up and initialized. Perhaps there is a reason it does not?

Also I think that the plugin should sense when there is no viable Polar.csv and default to "Computed from Sailboat Transform" (Weather Routing Config > Boat > Sail Plan Polar).

The weatherroutingconfiguration.xml looks like this now:

<?xml version="1.0" encoding="utf-8" ?>
<OpenCPNWeatherRoutingConfiguration version="1.1" creator="Opencpn Weather Routing plugin">
<Position Name="1" Latitude="29.91177" Longitude="-57.71040" />
<Position Name="2" Latitude="-38.81534" Longitude="45.00341" />
<Configuration Start="1" StartDate="7/31/2014" StartTime="5:55:00 AM" End="2" dt="3600" Boat="C:\ProgramData\opencpn\plugins\weather_routi ng\Boat.xml" Integrator="0" MaxDivertedCourse="90" MaxCourseAngle="180" MaxSearchAngle="120" MaxWindKnots="100" MaxSwellMeters="20" MaxLatitude="90" TackingTime="0" WindVSCurrent="0" AvoidCycloneTracks="0" CycloneMonths="1" CycloneDays="0" UseGrib="1" ClimatologyType="4" AllowDataDeficient="0" DetectLand="1" Currents="0" InvertedRegions="0" Anchoring="0" DegreeSteps="0.0;5.0;10.0;15.0;20.0;25.0;30.0;35.0 ;40.0;45.0;50.0;55.0;60.0;65.0;70.0;75.0;80.0;85.0 ;90.0;95.0;100.0;105.0;110.0;115.0;120.0;125.0;130 .0;135.0;140.0;145.0;150.0;155.0;160.0;165.0;170.0 ;175.0;180.0;185.0;190.0;195.0;200.0;205.0;210.0;2 15.0;220.0;225.0;230.0;235.0;240.0;245.0;250.0;255 .0;260.0;265.0;270.0;275.0;280.0;285.0;290.0;295.0 ;300.0;305.0;310.0;315.0;320.0;325.0;330.0;335.0;3 40.0;345.0;350.0;355.0;" />
<Configuration Start="1" StartDate="7/31/2014" StartTime="5:55:00 AM" End="2" dt="86400" Boat="C:\ProgramData\opencpn\plugins\weather_routi ng\Boat.xml" Integrator="0" MaxDivertedCourse="90" MaxCourseAngle="180" MaxSearchAngle="120" MaxWindKnots="100" MaxSwellMeters="20" MaxLatitude="90" TackingTime="0" WindVSCurrent="0" AvoidCycloneTracks="0" CycloneMonths="1" CycloneDays="0" UseGrib="1" ClimatologyType="4" AllowDataDeficient="0" DetectLand="1" Currents="0" InvertedRegions="0" Anchoring="0" DegreeSteps="0.0;5.0;10.0;15.0;20.0;25.0;30.0;35.0 ;40.0;45.0;50.0;55.0;60.0;65.0;70.0;75.0;80.0;85.0 ;90.0;95.0;100.0;105.0;110.0;115.0;120.0;125.0;130 .0;135.0;140.0;145.0;150.0;155.0;160.0;165.0;170.0 ;175.0;180.0;185.0;190.0;195.0;200.0;205.0;210.0;2 15.0;220.0;225.0;230.0;235.0;240.0;245.0;250.0;255 .0;260.0;265.0;270.0;275.0;280.0;285.0;290.0;295.0 ;300.0;305.0;310.0;315.0;320.0;325.0;330.0;335.0;3 40.0;345.0;350.0;355.0;" />
</OpenCPNWeatherRoutingConfiguration>

Editing Boat Configuration and saving to Boat.xml file

1. Start Opencpn, Climatology, Grib, Weather_Routing.
2. Highlight the route you created in the first menu.
3. Pick Configuration > Edit and "Weather Routing Configuration" opens.
4. Under "Boat" pick "Edit" this should be editing the "Boat.xml" file now.
5. Select "Boat Configuration" Tab.
6. Now edit the factors: In my case, I use Hull mono, Displacement 6, Sail Area 520, Lwl 22, Loa 32, Beam 10, for a SA Displ Ratio 15.87, Disp Length Ratio 503, Hull Speed 6.28
7. Capsize Risk 1.69, Comfort Factor 34.5
8. Pick "Save and Close".

Boat.xml now reads:

<?xml version="1.0" encoding="utf-8" ?>
<OpenCPNWeatherRoutingBoat version="1.1" creator="Opencpn Weather Routing plugin">
<BoatCharacteristics displacement_tons="6" sail_area_ft2="520" lwl_ft="22" loa_ft="32" beam_ft="10" />
<BoatDrag frictional_drag="0.0160" wake_drag="0.9000" />
<Plan Name="Initial Plan" polarmethod="1" eta="0.25" luff_angle="15.0000" wing_wing_running="0" optimize_tacking="0" />
</OpenCPNWeatherRoutingBoat>

I then went back and changed Displacement to 8 , picked "Save and Close" and then verified that it did change in the Boat.xml file. Then I changed it back to 6 and it changed back in the Boat.xml file.

Creating a Boat_Polar.csv file and using the file for a Routing.

1. Start Opencpn, Climatology, Grib, Weather_Routing.
2. Highlight the route you created in the first menu.
3. Pick Configuration > Edit and "Weather Routing Configuration" opens.
4. Under "Boat" pick "Edit" this should be editing the "Boat.xml" file now.
5. Select the "Boat Configuration" Tab which should show all the setting for your boat. Check it.
6. Select "Sail Plan Polar" the dropdown should show "Computed from Sailboat Transform".
7. Hit "Recompute"
8. Now hit "Save as CSV"
9. The file creation window pops up pointed at C:\ProgramData\opencpn\plugins\weather_routing and I entered "Nell_Polar_Computed.csv" and hit "Save".
10. Observing C:\ProgramData\opencpn\plugins\weather_routing a "Nell_Polar_Computed.csv" is created
    
11. The "Boat" menu dropdown still shows "Computed from Sailboat Transform" and is not referring to the new polar file that was created, so we have to select it.
12. From the "Boat" menu dropdown pick "Loaded from CSV File", then select the polar file you just created, in this case "Nell_Polar_Computed.csv" and Pick "Open".
13. Now there is a "Polar CSV File Settings > <Filename shown>" with the Wind speed step 3 and Wind degree step 3 and the dropdown shows "Loaded from CSV Polar File" Plot type selected is True Direction & Speed.
    
14. "save and close" the Boat menu.
15. Close weatherrroutingconfiguration menu.
16. Highlight your route "configuration" and rather than select Configuration > Compute pick Ctrl+C.
17. To recompute, you will need to first Configuration > Reset
18. Note that this routing is now using Nell_Polar_computed.csv
    
19. The plugin is quite powerful, each route "Configuration" will utilize unique settings for "Weather Routing Configuration" and "Boat configuration".

Thomas (Big_Speedy)
I think that if you can follow steps 1-10 at the top to set the "Boat" dropdown to "Computed from Sailboat Transform" you will be able to use the plugin without bombing. You should also be able to use your good boat polar file, provided it is adjusted to be the right format. --If you have it, just attach it here, and I'll try to adjust it so it works. What kind of boat is it?

Our computed and saved.csv Nell_Polar_Computed.csv is attached.