OpenCPN Beta Version 3.3.2330 Released

[cagney]

Originally Posted by MarkJ

I sail in pretty busy parts of the world... Currently where a "yacht" squeezing through a bridge may be worth more that $100 Million. I wonder if he would use degraded nav systems? Or be expecting me to?

You navigate the way you like. I likeColRegs that says navigate with ALL available methods. But as a reminder here is Section 7:

I think I know exactly what St Martin is like, after having spent a few years there, including umpteen times through both bridges.

It's certainly not a question of "degrading" a nav system, it's rather a question of warning users that they are using the system well beyond it's built in limitations.
I do expect a $100 Million yacht to have a captain that is familiar with all his navigation systems, including their limitations.

I once took a 48 foot cat through Sandy Ground, with a clearance of 1/2 a foot on either side. What do you think would have happened if I had spent time, viewing a screen, instead of the real world?

Thomas

[P_Dub]

Thanks for the reply I'll keep it shut in the future!
Bye

[MarkJ]

Here is the full screenshot of INavX.


So while you are berating the users you are designing a program when the "opposition" has a better program. They zoom in closer, clearly, than even your distorted, smudged one.

And you do it of your own volition.

Thats plain wierd.


Further, on a personal note, I have sailied in more than one country. No doubt many, many more than the developers, and many more oceans and ocean miles, more inshore and inland miles. So just pull up a bit on the reins before developers try to dictate to me.

Mark

[nohal]

Mark...
I will just repeat it again as it seems you missed it once more. You don't like it, you turn it off, you happy. Simple as that.

Pavel

[MarkJ]

And here is Garmins.

[bdbcat]

Mark...

Thanks for the comparison images. Supports our points exactly.

All of the charts use the same base cartography. Probably traced with a digitizer from a paper chart. Do you know where cm93 data comes from?

And the plain fact is that the features on this chart, or any overzoomed chart, may not be where they are charted, and it is difficult or impossible to estimate the inherent error. They will of probably be schematically correct in areas where people sail often, and may even be proven by recent survey to be extremely accurate.

But the provenance of the data is not carried in the database. Only the estimated geographical position. So OCPN, or any other system using the same root cartography, has no way to verify what it asserts about geography.

Paper charts in general are accurate to about 1 mm at say twice the stated scale. Nearly all vector cartography in common use comes from tracings of paper charts.

Those chart images you show are about 20x overzoom. This means that the accuracy is not likely better than 10 mm on screen. That's half an inch.

Now, you may usually navigate right down the middle of that channel with 2 meter GPS accuracy, assuming HDOP is 1 at the time you are transiting. These are your home waters, and you are comfortable there, maybe. Or maybe you have tweaked the cm93 offsets to create a good fit with reality.

But the real issue is navigating at a place you have not been before, in marginal conditions. This is where overzoom inaccuracy coupled with non-monitored GPS inaccuracy (time based due to HDoP variation) is eventually going to produce an ugly result.

And that's the point of overzoom emphasis in OCPN. To remind a user emphatically that the cartography is suspect.

I suppose you may have hard science, or pretty pictures, but not both in this case.

Anyway, I imagine we shall ultimately bring the overzoom emphasis disable control to the options screen. And there will no doubt be a mad rush for new users to disable overzoom emphasis immediately and permanently.

I will be sorry to see these users ignore reality for the sake of pretty charts, and even sorrier to see an accident that might have been avoided with all information, or lack thereof, in use by the navigator.

But some people don't wear seat belts in cars, either....

Dave

[colemj]

I don't understand this new-found concern for protecting the user. In OCPN I can turn off the display of navigation aids and use a font/detail level that is either illegible or impedes on navigation.

Why am I allowed to do this?

It is the responsibility of the user/navigator to understand the tools and use them properly. I won't accuse MarkJ of not knowing how to navigate and/or use navigation tools. I think anyone would have a difficult time making that case against me.

When OCPN succeeds in fully protecting all of us against ourselves, will you be removing the splash screen warning about using the program for navigation?

Mark

[NahanniV]

Instead of blurring the charts, would it be possible to blur the ship ?

What I am thinking of is to increase the size of the ship Icon based on the GPS accuracy and the Zoom level so that It would generally cover all it's possible positions.

Just a thought.

[rgleason]

Yes, I would suggest to just turn it off, if that is what you determine, while remembering Kiwi Vestas aground
Inside Track Special - Team Vestas Wind Runs Aground| Volvo Ocean Race 2014-2015

[dacust]

I think it is fine like it is.

Myself, I would probably turn it off, but then I've done a bit of research into chart errors and dangers and think I am OK without a seatbelt,, uh,, I mean chart blurring.

But, here are two other ideas just to add fuel to the fire since it is so boring in here:

(1) An old Garmin I had plotted a circle for the predicted accuracy based on current reception. You could put fine line hash marks inside the circle so when you get to 20x or bigger, maybe the hash marks cover the entire screen.

(2) After zooming to max scale, treat it like a raster. Just take the max zoom image and enlarge it. Let the lines, words, images, etc. all get larger together.

(3) both

-dan

[bdbcat]

Gilletarom...

re:
"I get exactly the same views with PolarWiew as with OpenCPN. Here is a screenshot of PolarWiew. There is a list of all the charts of the set charts for the area requested by Dave.

Please Pavel, what is your chart ?"

You should download District 1 RNCs.
http://www.charts.noaa.gov/RNCs/01CGD_RNCs.zip

The largest scale chart at the location we are considering is

12364_18.KAP

This chart should be in that set.

Thanks
Dave

[NahanniV]

Originally Posted by Hakan

Dave and Caesar...
I'm sorry, the font change was introduced in 3.3.2316 and not in this beta.
I havent' had this toolbox open since then. I've tried a couple of beta updates and the change is after a pull I built 2014-11-24. The first picture here is from that build, and it's not spaces as I wrote.
When I run from a built as of 2014-11-26 the font change is introduced and the squares are there. See next picture.
BR24Radar_pi isn't changed during this period.
Maybe you can find out what commits are made then?

Håkan

I have been testing the br24 plugin during the same time frame and have not had a problem with fonts.

I built it today for ARM and ran it with latest OpenCPN.

Also running on Win7.

Cheers,
JM.

[bdbcat]

Håkan...

On the button font problem:

Good news and bad.

1. Good news. It can be fixed.
2. Bad news. The fix should be in the PlugIn, not in the base.

Turns out there is a bug in wxWidgets 2.8. It shows shows up on XP only, I think.

Info here:
#12491 (Multi-line wxButton labels) – wxWidgets

Multi-line button controls need to be initialized with their ultimate line count if you plan to use the default system color for the control. Our recent change to OCPN base does just that. We use the system default colors in RGB and DAY colour modes.

You can fix the PlugIn simply, at least for some simple buttons, like this:

in br24radar_pi.h...

    RadarControlButton(wxWindow *parent,
                       wxWindowID id,
                       const wxString& label,
                       br24radar_pi *ppi,
                       ControlType ct,
                       bool newHasAuto,
                       int newValue
                      )
    {
        Create(parent, id, label + _T("\n"), wxDefaultPosition, g_buttonSize, 0, wxDefaultValidator, label);

Note the extra "\n" in the Create() method. Silly, I know, but that's the bug. If we try to fix it in the core, we break a lot of other things.
So....

Good Luck
Dave

[RhythmDoctor]

Dave - Thanks for adding the Toggle Full Screen option to the context menu. As you know, I use tablets a lot during sailing season, but I also do planning and testing on desktop and laptop systems at home. Sometimes my keyboard is in an inconvenient place, and mouse commands are easier. For this reason, I'd like to double-down on my request to have a Full Screen Toggle available always via mouse command, no just when Tablet Mode is enabled.

I suspect that you want to minimize the length of the context menu whenever possible. So I'd be happy with having the toggle command somewhere else - either a checkbox on display options, or a toolbar icon (hidden by default to minimize clutter). But I do think that it should be available. I am already using my desktop machine in tablet mode just so I can toggle full screen with my mouse. I would prefer to have this capability available without being in tablet mode.

FWIW, I still think blurring/fogging is a good idea. It is actually very dangerous to have the chart rendering implying greater precision than the underlying data, and O needs to at least default to a degraded image as a warning. (I guess people can turn it off if they choose to.) I've heard many complaints from navigation instructors about the dangers of overzooming, and I think it's good for O to be at the forefront of this important safety issue, even if a few people insist on turning it off through the .ini file, or by turning off quilting.

[Hakan]

Originally Posted by bdbcat

Håkan...

On the button font problem:

Good news and bad.

1. Good news. It can be fixed.
2. Bad news. The fix should be in the PlugIn, not in the base.

Turns out there is a bug in wxWidgets 2.8. It shows shows up on XP only, I think.

Info here:
#12491 (Multi-line wxButton labels) – wxWidgets
....................Good Luck
Dave

Dave..
Many thanks for your investigation.
I've tested it on Win7 and got the same display there as well.

I'll try your suggestions in my copy and test. Hopefully it works nice and I push that to Kees original source.

Håkan