Celestial Navigation Plugin Redux

[rgleason]

Just saw Seans note. I guess the plugin is in good hands. Stelian and Sean thankyou.

[P_Dub]

One of the only thing that I am not finding to work correctly is moving the position line to a subsequent sight (or sights) to the next DR later in the day to make a fix. Say sun sight at 09h00 and noon sun sight. Anyone can explain what I am missing in that system.
Bye

[rgleason]

Just cloned Stelian42/cel_nav and got some compiler errors c2027

https://stackoverflow.com/questions/...-declare-class

https://stackoverflow.com/questions/...nd-error-c2227

https://msdn.microsoft.com/en-us/library/6c2dk0ah.aspx

Are there any binary's needed? Don't think so. See attached please.

Will see if you manage to compile it under "Releases"

[rgleason]

Here are some of the lines and error codes... sounds like it is a format error. Windows is picky unfortunately.

...\src\FindBodyDialog.cpp(45): error C2027: use of undefined type 'wxFileConfig'
double lat, lon;
celestial_navigation_pi_BoatPos(lat, lon);
wxFileConfig *pConf = GetOCPNConfigObject();
pConf->SetPath( _T("/PlugIns/CelestialNavigation/FindBody") ); <---line 45

...\src\FindBodyDialog.cpp(45): error C2227: left of '->SetPath' must point to class/struct/union/generic type
double lat, lon;
celestial_navigation_pi_BoatPos(lat, lon);
wxFileConfig *pConf = GetOCPNConfigObject();
pConf->SetPath( _T("/PlugIns/CelestialNavigation/FindBody") ); <---line 45

...\src\FindBodyDialog.cpp(48): error C2027: use of undefined type 'wxFileConfig'
bool boat_position;
pConf->Read( _T("BoatPosition"), &boat_position, true ); <--- line 48
m_cbBoatPosition->SetValue(boat_position);

...\src\FindBodyDialog.cpp(48): err or C2227: left of '->Read' must point to class/struct/union/generic type
bool boat_position;
pConf->Read( _T("BoatPosition"), &boat_position, true ); <--- line 48
m_cbBoatPosition->SetValue(boat_position);

...src\FindBodyDialog.cpp(52): error C2027: use of undefined type 'wxFileConfig'

..\src\FindBodyDialog.cpp(52): error C2227: left of '->Read' must point to class/struct/union/generic type

Also Lines, 56, 57, 70, 71, 72, 75 and 77. Similar.
====

C2227 Errors

https://social.msdn.microsoft.com/Fo...orum=vcgeneral

https://stackoverflow.com/questions/...te-class-error


====
p_dub are you Linux OS?

[rgleason]

Sean's version gets the same errors!

[stelian]

Originally Posted by rgleason

Sean's version gets the same errors!

Yeah, same here. Builds fine under Linux but not under Windows. It comes from Sean's changes.

[seandepagnier]

Originally Posted by P_Dub

One of the only thing that I am not finding to work correctly is moving the position line to a subsequent sight (or sights) to the next DR later in the day to make a fix. Say sun sight at 09h00 and noon sun sight. Anyone can explain what I am missing in that system.
Bye

You are supposed to first take a sight at 9 am.

At 12 am, take another sight, and edit the 9am sight adding a bearing and offset. Do not change the time of the 9 am sight, it will always be at 9 am.

[stelian]

Originally Posted by stelian

Yeah, same here. Builds fine under Linux but not under Windows. It comes from Sean's changes.

Sean fixed the build error, I confirm it builds ok on Windows now.

I've merged Sean's master into my repo, ready to be pulled !

[P_Dub]

====
p_dub are you Linux OS?[/QUOTE]

No, I am windows 10.
I don't have anymore Visual Studio, so can not compile.
If you can send me a new one compile.
Like I said before, we need to put in the interface a DR position in order to be able to discriminate between the 2 positions of circles crossing each other ( in the case of 2 sights) and also, when 2 sights are hours apart, like morning sun and noon sun. that is the normal procedure at sea. I know it may not be easy but if someone could do it this would nice to have for you fellows.
In the means times I keep checking.
Regards.
P_Dub

[rgleason]

Thanks P_Dub, Will try to remember you are windows.
Stelian I hope to compile cel_nav next. Have been working on weatherfax and learning how to use git better today. Cel_Nav next

Sean showed me that we have that feature and explained that When you do a FIX you need to know very roughly which X is the right one. This is done by putting just integral numbers for LAT LONG in bottom PART of the FIX Menu. I think we need some title that helps with this. Think about it.
Only integers are needed, very rough for my example 42 -71 would work.

Like I said before, we need to put in the interface a DR position in order to be able to discriminate between the 2 positions of circles crossing each other ( in the case of 2 sights) and also, when 2 sights are hours apart, like morning sun and noon sun. that is the normal procedure at sea. I know it may not be easy but if someone could do it this would nice to have for you fellows. In the means times I keep checking.

[rgleason]

P_Dub, you'll find a compiled file here.
https://github.com/rgleason/celestia...es/tag/v2.2003

Its much better. Make sure UT is correct before using FIND.
I still have 53nm off from Boston 42.35 -71.1, at the same time 1800 Oct 10, 2017

...but the sights are crossing much much better.

See attached files and screenshot.

Thanks Sean and Stelian! It is much better and easier to use. Boat Pos is working in FIND and I really appreciate the copy measurement feature in FIND!


Looking much better. P_Dub give it an exactness test please.

[rgleason]

The crossings of the sights seem closer, but they're still about 57nm away from the VDR file I made for Boat Position of 42.35 -71.1

I think the Clock Offset is working great. I used it to find out that the sights are -318 seconds off. I wonder why they would be off that much? -57 nm = -318 seconds I guess, but where would that come from?

The FIX just needs approximate Full degrees Long & Lat to find the correct crossing (there are just two for two circles for example) Sean says the Sphere and two cones are just different ways of calculating the fix. You should not see a big difference in the fix or something is wrong.
What should the description say here?

Rough Lat ___ Long ___ use Full degrees?

[rgleason]

I found two time errors that did not seem to help too much, then added more sights.
These sights seem to be east more, but the FiX is further off.

Sights and screenshot are attached.

[rgleason]

Sean's responses quoted:

The crossings of the sights seem closer, but they're still 54nm away from the VDR file I made for 42.35 -71.1

Yes, well the find body is only approximate. It doesn't perform the atmospheric corrections and height of eye and other calculations, this is why they don't align perfectly. When calculations are performed the other way around for real sights, all the corrections are performed.

It might be nice to improve the find body to perform all the corrections in reverse, but it does explain why other people have verified the plugin does work for real sights.

I think your Clock Offset is working great. I used it to find out that the sights are -318 seconds off. I wonder why they would be off that much? -58 nm = -318 seconds I guess, but where would that come from?

Again, this is not the issue. I still didn't make the clock offset update in realtime.

[kubek]

Hi,

I can confirm that "Beta celestial_navigation_pi-2.2.003-ov4.-win32" version is accurate.
I put my 5 physical sights from summer voyage Canaries - Cabo Verde and accuracy is less then 2 NM, only one sight have near 3 NM. So plugin works perfect and those errors are like human measure error.

Great work!

Thanks again,

Kubek