OpenCPN Version 2.1.0 Beta

[balp]

Is no longer working with Ubuntu 10.04, the version for libgps miss match. Is there any great stuff we now need?

[balp]

Added a patch that works for me.

https://sourceforge.net/tracker/?fun...42&atid=894807

[Netsurfer]

Dave,

it seems that there is a fault in the calculation (or display) of the Look Ahead in Course Up mode feature (see screenshots).



As you can see the COG is 270° whereas the look ahead is like cruising 180° (90° error).
To my observations this happens in E <-> W directions (same on course 90°).
N <-> S seems to be laterally transposed (couldn't test directions between yet). So when COG is 0° the ship is positioned at the upper border of the chart window and vice versa.



Gunther

PS: Simulation with 9600 bps works fine.

[cagney]

Dave

File "VERSION.cmake" has not been updated to 604 in CVS.

Thomas

[nohal]

Dave,
I found an ENC cell which crashes even the latest CVS code while ingesting. Opens normally in SeeMyEnc, nothing special in the log.

Pavel

[cagney]

I find the same thing as netsurfer regarding lookahead.
Testing with the AIS simulator, I find that it works more like a "look behind" feature, perhaps just a sign error somewhere in the calculations. I have not tested on an E-W course.

And....Thanks Dave for another great beta release!

Thomas

[idpnd]

Originally Posted by cagney

And....Thanks Dave for another great beta release!

Yes RC as well.. Press release for 2.1.0 might be in order

Weird peak in downloads just now, must be an sf.net error?

2010-05-29 147
2010-05-30 213
2010-05-31 199
2010-06-01 53,765
2010-06-02 91,620
2010-06-03 163
2010-06-04 125
2010-06-05 22

[Netsurfer]

Originally Posted by cagney

I find the same thing as netsurfer regarding lookahead.
Testing with the AIS simulator, I find that it works more like a "look behind" feature, perhaps just a sign error somewhere in the calculations. I have not tested on an E-W course.

Addition:
On COG 180° it works as expected (but only on bearing South) ...

And....Thanks Dave for another great beta release!

and the new features!

Gunther

[cagney]

No, not a bug report, just a few thing to think about before the next stable release.
Picture below shows a quilt between a gnomonic projection chart, to the left, that I have treated as a mercator chart, and a mercator chart to the right.
Gnomonic charts were BA standard for harbour plans and other large scale charts before a transition to tmercs started. A few of these are still around. If I had labeled the chart as "gnomonic" I suppose it would have been excluded from the quilt, but still rendered as a merc - Is this correct?
I just read the other day that most charts labeled as "Gnomonic" by BA actually were polyconic.(can't remember where I read it.)
I'm not sure we need to do much about these charts, but they do exist in the "wild".
Talking about Transverse Mercator charts (and UTM), I can't see any advantage for the navigator in this projection. I suppose it must be convenient for the chart maker to produce these charts. Anyone with any insights in this matter?

Toolbox => Own Ship Displayoption => Arrow length in minutes. I think that this option now defaults to 5 minutes. I suggest that this is changed to 6 minutes for easy calculation. The arrow distance will always be 1/10 of the distance that own ship covers in an hour. If you are doing 9 knots the arrow will be 0.9 miles long etc.
Of-course anyone can change this to their liking......just thinking about sensible defaults.

Thomas

[cagney]

Originally Posted by balp

Added a patch that works for me.

https://sourceforge.net/tracker/?fun...42&atid=894807

Dave,
As balp found out, Network libgps is not displayed as a choice for NMEA Data Source, and gps on Linux with gpsd 2.9.4 is no longer working.
Unfortunately balps patch did not work for me.

Thomas

[manimaul]

@Thomas

I can't believer how fast you update the download page on opencpn.org including an rpm and deb for every build. You're an animal on all the documentation pages you've written in the wiki too. Amazing!, thanks.

-----------------------

@Gunther

Originally Posted by Netsurfer

Addition:
On COG 180° it works as expected (but only on bearing South) ...

and the new features!

Gunther

The compass always points North (Not course) ... With that realization in mind, I think you can see that it is working properly.

Will

[bdbcat]

nohal/pavel....

Sorry, I cannot see which ENC file is failing. Can you give me the filename/source?

Thanks
Dave

[bdbcat]

Cagney/thomas....

Did you get any dialog regarding lbgps version mismatch on the failure to find gpsd ? Did the cmake output indicate the the the libgps headers were found?

Thanks
Dave

p.s. looking at lookahead bug now. Obviously I did not do enough alpha testing....

Anybody else able to test 9600 baud with real hardware? I can't

Thanks
Dave

[cagney]

Dave

Cagney/thomas....

Did you get any dialog regarding lbgps version mismatch on the failure to find gpsd ? Did the cmake output indicate the the the libgps headers were found?

Nada! I cleaned out everything and started a clean new download and installation and tee:ed all output from my installation script to a file. Not a word about gpsd!!

In CMakeCache.txt I found

//Enable GPSD Library
USE_GPSD:BOOL=OFF

I changed this to "ON" and ran "make" again.
I then saw this output...

-- GPSD Library support: enabled
-- Found gps: /usr/lib/libgps.so
-- Building with libgps includes

Trying to run opencpn this time produced the dialog you asked for. Applying balps patch, recompiling again, and ...things just worked.

I haven't had time (yet) to check what caused the problem in the first place. It would be useful with a warning, if gpsd support was not enabled.

Thomas

[bdbcat]

Thomas....

So, the cmake scipt did not find the libgps includes on its own.

The logic goes like this:

1. If cmake finds libgps include files, then opencpn will be built with libgps support.
2. if built with libgps support, then the entry "libgps" will be in the NMEA selector control, whether or not a usable libgps exists on the target system.
3. If the user selects libgps as NMEA source at run-time, and a usable libgps library is found (Balps version test patch needs to be included..), then all will be well. If the library is not found, or too old, a message is presented.

So, the question is: I did not change cmake script for Beta 604. Why did cmake fail to find the libgps headers this time?

Aha! I just thought of something. cmake is not reliable about its cache management. The search for libgps (and lots of other things)is cached from run-to-run. Mistakes are often made.....

I find it best to

$rm -f CMakeCache.txt

between cmake runs, especially if I have changed something.....

cmake has no built-in cli option to force a new cache. It can be done from windows gui. "Delete Cache" button is right there.

What were they (not) thinking?

Wondering....
Dave