OpenCPN Version 2.1.0 Beta

[yachtvalhalla]

Originally Posted by PjotrC

In ais2.jpg the vessel selected (Tanker ARCHIMD) is broadcasting AIS status "underway", and is displayed accordingly. Do you have a case where navigational status does not match display?

I think you mean ais1.jpg for ARCHIMD. Look at ais2.jpg (HHIC-F1). It doesn't have an entry for Navigational Status. And, as shown in the image nuc.jpg (from another AIS display program) here's a vessel showing NOT UNDER COMMAND!.

So here is the problem: OpenCPN is using the Navigational Status to determine if a vessel is anchored. This is not going to work. The static data is entered by the crew and is more often wrong than right. That's why other programs use the SOG as the determinant, recognizing that a GPS at rest may still show some low boat speed. A threshold of 0.3 knots usually works but I find that 0.5 is a better determinant. Since OpenCPN has the ability to specify 'Suppress anchored/moored targets' based on target speed, I think this parameter should be used for the display of anchored vessels (with a black ball).

Just my 2 centavos.

Terry

[bdbcat]

Team....

OpenCPN Version 2.1.0 Beta Build 331 is now available for download.


http://www.southbaynetwork.com/openc..._210_setup.exe

Changelog:

30. Improve quilting algorithm and WVS background for alternate chart types, esp cm93.
31. Correct cm93 offsets for certain cells having multiple M_COVR regions.
32. Correct fault when deleting an active or current track.
33. Add RTOFS Current and Sea Temperature data to GRIB display.
34. Allow '=' key as substitute for '+' for Zoom-in.
35. Symbolize moored/anchored AIS targets with black ball.
36. Correct initial chart open scale to use last saved scale from config file.
37. Reduce cursor track timer interval for __WXGTK__ builds to 2 msec.
38. Add Russian and Czech language support.
38, Correct crash on Track Delete with Track Properties dialog open.
39. Allow adjacent/coincident TrackPoint load from gpx file.

----Release Beta 2.1.0 Build 331----

This Build fixes several of the reported problems so far.

We need more feedback on unexplained crashes on startup under Windows. I have corrected nothing obvious in this area, as I cannot reproduce....

I am tracking this thread closely.

CVS is __NOT__ updated as of 0200 Z 01April. CVS servers are down for maintenance. I will sync ASAP.

Thanks, and good luck
Dave

[yachtvalhalla]

Thanks for the updated build. (Yea! BLACK BALLS)

I'm still convinced that the SOG should be used to determine the anchored or moored status ILO Navigation Status.

Re aistargetlist.jpg: (Taken from Yacht-AIS Pro and would be a nice feature to have in OpenCPN ... sorry for the OT).

The static data entered by the crew cannot be relied upon. In this case (red ellipses) four vessels have not changed their NavState. Eleven vessels have not deleted their ETA on arrival. One joker in the crowd .. destination of "GROUNDHOG DAY". The crew cannot change the dynamic data (position, SOG, COG, etc.) so it remains as the only reliable determinant for the stationary status of the vessel.

Re ais6.jpg: (From OpenCPN beta) Three of the four improperly designated NavStates (from above) should be shown with the green/black ball symbol. Don't know why the fourth one isn't shown ..

Sound reasonable?

Terry

[Totobeloeil]

Fresh install on Windows XP (no previous version of ocpn) : got an "info" window then an "error" window at first launch

When installing v2.1.0 upon v1.3.6 on Windows XP, no problem at first launch.

Attached files are from fresh install test.

Jean-Pascal

[Viking Sailor]

Dave,

A problem with build 331 on XP home when trying to do CM93 quilting.

The following show the results of selecting CM93 quilting options.

Paul

[surfjoe]

Hi Dave,

sorry, but I have the same problems as stated in my post #9:
1. quisting with CM93

Attachment 14676

2. Symbols for different buyos still missing
Is it helpfull for you to have a list of the different CM93-Objects? I found a file CM93ATTR.DES in the program folder of my navi-SW. Therin you can find all objectcategories with there different characteristics.

Best regards and thanks a lot from GER
Joe

[jan eulen]

Hello Dave, I found a bug I think.
I run windows xp. To move around I use the up, down, left and right arrow. The down arrow does a zoom-in action and no moving around. The other arrows work fine.

Thanks for your work regarding opencpn.

Jan Eulen

[LeaseOnLife]

Dave

33: Nice! For anybody cruising across or along the gulf stream, this is very valuable.

One suggestion: Could the data display for the grib be limited to the data included in the grib file? For example, if no significant height wave data are in the grib, don't show an empty box for wave height. Or, when displaying an RTOFS GRIB, don't show 4 empty boxes for wind speed, direction, pressure, significant wave height. See screenshot

One bug I found when displaying RTOFS GRIBS: when the cursor is outside the GRIB area, the sea surface temperature box should be empty, but it shows some weird value: "-1000000272.15"

As a feature request (old one): Could you also add the "RAIN" parameter to the GRIB file display? For example overlay a black "sprinkles", with increasing density for higher RAIN numbers.

[LeaseOnLife]

+1 on this bug. The down cursor does pan the chart, but it also zooms in

Originally Posted by jan eulen

Hello Dave, I found a bug I think.
I run windows xp. To move around I use the up, down, left and right arrow. The down arrow does a zoom-in action and no moving around. The other arrows work fine.

Thanks for your work regarding opencpn.

Jan Eulen

[Netsurfer]

Dave,

Originally Posted by bdbcat

If I understand correctly, the last opencpn.ini file created by Version 1.3.6 under Win7 causes 2.1.0 328 to crash immediately on the first run after installation. Is this so?

That's correct.

If so, could you recreate the problem by re-installing 1.3.6, running it for a bit, and then installing 2.1.0 and verifying the crash? Then, if possible, please send me the opencpn.ini which causes 2.1.0 328 to crash, and I will investigate with debugger.

Kind of fiddly work, but I will appreciate it...

Thanks for the offer ..., I tried to reproduce the crash, but couldn't make it.
Maybe (and really just maybe) I have tried to install the new version beside at the old one in a new directory.

Due to OCPN creating only one INI-file with the path to the installation directory, multiple installations are not quite functional. Therefore I would recommend some changes/ additions to the installer (at least for the Windows distribution):

1. The installer should look up for an existing version and if found offer an upgrade option (such as is usual for Browser like FF)
2. As long as there are no different options to choose from the step of selecting components should be skipped (just make no sense to offer a selection when there actually is nothing to choose from).
3. The uninstall routine should offer the possibility to also remove/ delete the settings dir (which is somewhere at .../AppData/Roaming/opencpn). Non experienced users (and that is the majority) won't be able to delete their old settings otherwise and could never start a clean installation from scratch.

Another option worth thinking about is to change the program settings part to make multiple installations possible. Not having them working simultaneously. This would be especially helpful for users using OCPN for real navigation. So they can use the latest version with all its new features and switch back to the old (at least more stable) one in case a bug appears or even worth at program crashes.

I'm a bit wondering that nobody else seems to have the zoom problem!?
For me it is still present. I checked back to version 1.3.6 and the (or one) difference is that in version 1.3.6 I can only zoom out to level 0.42x but in 2.1.0 to 0.25x. And that seems to be the problem. As soon as the zoom level is smaller than 0.5x (next step is 0.25x) the problem is present.
Can anybody confirm this?

Gunther

[LeaseOnLife]

I am playing with PiotrC's anchor watch. Piotr, thanks so much for this feature.

A couple of things I would like to see improved:

• I think any waypoint should be able to be the anchor watch center (like: right click on waypoint, "use for anchor watch"
• The alarm distance should not be "hidden" in the name of the waypoint but be a generic value to be set in opencpn, via config or rightclick menu
• I don't get any sound in the event of an anchor alarm. AIS alarm works (Windows XP)
• This goes beyond the anchor watch feature: At close up scales, how about displaying the boat at real scale, including the position of the GPS sensor Here is a good example: The screenshot shows "NavMonPc" doing anchor-watch. The center, is where I think the anchor is. The green circle is the anchor watch distance, set to 120ft. My GPS sensor is at the port stern. A visualization similar to this, on top of a chart, is what I am looking for.
• Then, Piotr already picked up on this, it would be great to be able to color tracks by age, when swinging in places with tidal currents, it would be good to set the aging of tracks to 1/2 or 1 tidal cycle.
• Sometimes one doesn't want to swing full circle. See screenshot two. I would like to be able to define a sector as a danger area. Like when the trade winds die and I suddenly drift towards the reef.

[idpnd]

You can implement the "danger area", you can do a proximity alarm with a negative number. The gps sensor location has already been discussed, I think its a bit over the top personally and I do recall that this was the general consensus.

[PjotrC]

Originally Posted by yachtvalhalla

Thanks for the updated build. (Yea! BLACK BALLS)

I'm still convinced that the SOG should be used to determine the anchored or moored status ILO Navigation Status.

Sound reasonable?

Terry

Sorry for the delay, I was travelling most of the day, and the boat is now in drydock, so lots of action...

Yes, what you write is reasonable. I know the facts from experience, and I know the widely read work from Liverpool University on Human Factors in AIS. The official AIS navigational status and practical description of current maneuvering mode can be two different things. One should be interested in both.

AIS status "At Anchor/Moored" is exactly what the black ball says. The ball is also hoisted by the crew, falsely or not. It is a declaration only, and we know this. To fully assess the situation one looks at the ball, and one looks at the anchor chain at their bows, and at prop wash at the stern.

One of the official speed threshholds for anchored/moored ships is 3 knots (the transmitting interval changes then from 3 minutes to 10 seconds, but it does not mean the ship is not at anchor then).

Probably a different terminology should be used. The option many people want to have implemented should be called "remove stationary and slow moving targets". The relevant targets could be flagged in some special way as "stationary or slow moving" (but not black ball). In fact, they are - by the cog/sog predictor vector. Nothing wrong in that.

The AIS definition is clear - we know what parameters are set by fallible humans, what - by differently fallible hardware, and we make our judgements. I just propose to call things as they should be called. There is no other way to determine if a ship is "at anchor or moored".

Re: possible bugs in displaying status (other than the above) - I'm afraid that I need AIS data to debug. I will be trying to monitor local AIS, and please record and post some AIS NMEA that works improperly. This surely needs fixing.

[PjotrC]

Originally Posted by LeaseOnLife

A couple of things I would like to see improved

Thanks for the suggestions. In fact, it was also your post that inspired me to add this little feature...

1. Any waypoint for anchorwatch - I think one can use any mark/waypoint that is close enough to the boat. The reason is that it could be confusing if one e.g. by mistake sets a watch on a WP 100 miles away (this might be invisible, hard to find, etc.). Also, I think more than 2 could be unmanageable and too complex to average navigator.

2. The method of entering distance. Yes, the usual way would be a special dialog or menu - I wanted to keep the user interface as thin as possible. It is simple and it works, and is displayed conveniently. Does one often set anchor watch on a RouteWaypoint or NavAid?

3. The sound. When implementing the prototype, I did not find an (easy) way to make two kinds of alarms (AIS and anchorwatch) coexist, at least without making changes to OpenCPN structure in-the-large. I'm sure sound will come when the accepted functionality will stabilize. It might be also time to decide if we want still more alarms (XTE, depth, COG, whatever...).

The more advanced suggestions... as you say, they go beyond the starting minimum. Let's get the first version stable...

[cagney]

The CVS server is back online!