OpenCPN Version 2.4 Beta Build 310

[bdbcat]

Ahoy OpenCPN'ers...

OpenCPN Version 2.4 Beta Build 310 will soon be available for download from opencpn.org.

Monitor the opencpn.org Downloads page to find and download the latest Beta version.....

Source is available now on sourceforge git repo.

Changelog:

1. Implement support for BSB Version 1 Polyconic projection charts.(FS#326)
2. Implement cm93 "per-cell coverage" offset adjustment dialog.
3. Implement Transverse Mercator (TMerc) quilting. (FS#315)
4. Implement Full screen quilting mode, Toolbox selectable.
5. Implement Blit accelerated chart panning.
6. Improve overzoomed quilt patch registration and georeferencing.
7. Correct some Route and Waypoint navobj regression errors. (FS#244, 324, 327, 343, and 349) Thanks for the patch,Pavel.
8. Implement configurable Lat/Lon grid lines. Thanks for the patch, Carsten.
9. Correct detection of some BSB charts in UTM projection (FS#325)
10. Implement Multi-Dashboard support. (FS#289) Thanks Sethdart
11. Improve Dashboard NMEA messsage parsing for Wind and Heading instruments (FS#276) Sethdart
12. Add additional translatable strings to AIS message parser. (FS#351) Thanks again Sethdart
13. Various language updates (ET)
14. Implement PlugIn API Version 104

Release Beta Version 2.4.310

Notes:

2. CM93 "per cell" offset dialog.
Background:
Each cm93 cell contains one or more MCOVR objects, describing the datum and possible offsets which apply to objects within the boundaries of that MCOVR region. What is new here is that we have implemented programmable "user offsets" for each MCOVR object as well, so that individual coverage regions may be fine tuned by users so that they may better align with observed georgraphy or other (presumably more accurate) charts. User offsets will apply only to the MCOVR region(s) selected, and not to the entire folio of cm93 charts as is done in Version 2.3.1.

Here is how to use this feature.
1. Go to single chart (i.e. non-quilted) mode, with cm93 charts displayed.
2. Zoom/Pan to the region of interest.
3. Right click, activate "CM93 Offset Dialog"
4. In the list of cells/MCOVR IDs, selected the one MCOVR of interest. The coverage area for this MCOVR object will be outlined
with a heavy yellow border.
5. Use the spin controls on the right side of the dialog to adjust the desired user offsets.
6. The offsets are automatically saved in the binary MCOVR cache files found in the program data directory as CM93/...

In order to use this feature effectively, one needs a geographically known reference point. As an example, say there is a NAVAID present on the chart cell with known wgs84 lat/lon, as from a light list. In this case, one can drop a waypoint on the chart, edit the waypoint properties to the known lat/lon, and then use the CM93 Offset Dialog to slew the cell/MCOVR so that the dropped waypoint coincides with the NAVAID feature on the cm93 cell (chart).

You may also use a geographical feature such as a cape or point whose location is absolutely known, as by reference to observation, or by rendering on another, more accurate chart such as a trusted scanned raster chart or georeferenced photo chart.

3. TMerc charts are quilted with other TMerc charts. This should work well in areas (e.g. Sweden) where there are multiple overlapping
North-up TMerc charts. It will not work in South Pacific, where there seem to be isolated (NZ?) TMerc charts.

4. By request, we allow more charts on the viewport to be added to the quilt. We add charts of an appropriate scale range to the quilt, even if they do not overlap with the center of the screen. Of course, this will cause some performance issues for some locations and some systems. So, it is configurable in Toolbox.

14. PlugIn API Version 104: This will be transparent, but will enable the BSB4 and other PlugIns coming soon.

Good Luck testing, and keep those Flyspray tasks coming.
Don't forget to update "your" Flyspray tasks with with test results. That way we can close the tasks as appropriate.

Sorry about the temporary un-availability of the Windows installer. Should be uploaded soon.

Thanks
Dave

[janp391]

Thanks for the beta
I am running XP home SP3
The beta and new plugin installed nicely. But when zooming in on cm93_2 some , not all, charts never came down in detail level where all nav objects was visible.
Uninstalled 2.4.0 and went back to 2.3.1. and all was ok again.

[nohal]

Dave,
thanks a lot for the beta. Generally pretty stable
First observations:
It looks like OpenCPN does not react properly on changes in the INI file done outside of OpenCPN.
The scenario:

• OpenCPN installed with charts (US RNC + CM93, but probably it's not important)
• INI deleted while OpenCPN was not running
• After new start the charts are still there, while of course no chart dirs are defined by the config.

I suppose that on startup the CHRTLIST.DAT content not meeting the ini settings should be discarded.


I can't zoom out above certain level. For example looking at US RNC 11520_1.KAP and quilting off I can't get above TrueScale 503900 / Zoom 0.73. I think it's the same thing discussed in some other thread where cagney didn't have this problem while others did, but somehow can't find it right now...


And the last one: artifacts stay in the areas without charts while panning the quilted charts with mouse. Screenshot attached.

Pavel

[sailias]

Build from source on Ubuntu 10.10. Compiles without errors and work swimmingly. Thanks so much for all the hard work.

/ch

[CarCode]

Comment to note #2 above (CM93):

If you extract the cell numbers and perhaps also the Lat/Lon of each cell why not make a chart manager of them like the chart manager in PolarView?
BTW, is the leading zero of the cell numbers suppressed in the new offset tool?

Most people installs the whole world CM93 package and does not sail around the whole world. A chart manager might help to delete the charts not needed or to activate others which is otherwise rather complicated now for CM93.


A real good chart manager should have a high priority in my point of view.

Gerhard

[cagney]

Dave

Thank you for another very interesting beta.
I'm in the process of testing different features and will enter some of my findings in "Flyspray".
I have found a new 2.4 crash when zooming out in quilting mode using CM93. There is no limit on zooming out anymore. You can go on until it looks like this:


If you then zoom out a bit further, until the world is a black dot and then try some panning, a repeatable crash occurs.

Thomas

[bdbcat]

janp391....

Thanks for your report.

Are you experiencing a problem in cm93 quilt mode, or single-chart mode?
Can you provide a screenshot of the problem, or give instructions to reproduce reliably?

Thanks
Dave

[janp391]

Originally Posted by bdbcat

janp391....

Thanks for your report.

Are you experiencing a problem in cm93 quilt mode, or single-chart mode?
Can you provide a screenshot of the problem, or give instructions to reproduce reliably?

Thanks
Dave

Thanks for reply
I am not in quilt mode, just moving around on the chart and at some charts all details is gone like as one zoom level on one chart and another on the other chart.

[CarCode]

Originally Posted by janp391

Thanks for reply
I am not in quilt mode, just moving around on the chart and at some charts all details is gone like as one zoom level on one chart and another on the other chart.

But I reproduce your screenshot only when in quilt mode. Please shut off quilt mode.

[cagney]

janp391
Your screenshot shows that you are using quilting.
Have you ticked the new switch, ToolBox->settings->"Disable Full Screen Quilting" ? If yes, turn this switch off and I would guess that you'll have the old behavior back again.
Thomas

[Patg]

Hi folks

For the brave ones, MacOSX Alpha of 2.4 Build 310 is at

ftp://ftp.imp.ch/pub/opencpn/opencpn...0_20110312.dmg

Regards

Patrick

[bdbcat]

janp391....

Thanks for the screenshot.

Still having trouble reproducing this effect.
With respect to the your attached screenshot, this is shown in quilt mode. OK, no problem. So, what is the setting of the cm93 detail slider in Toolbox->Vector charts?

I observe:
The left portion of the shot is cm93 EScale, and the right is cm93 CScale. Somehow, the screen shot is missing the available DScale chart for the portion on the right side.

But I cannot reproduce. Can anyone else?

Position is 58 05N 011 30E

Thanks
Dave

[cagney]

Dave
I can reproduce it, kind of. See dumps below. Detail slider = 3. The new setting mentioned in my last post made no difference, if on or off in this case.

With quilt
Without

I'm using the CM93 version from February 2010.

Thomas

[janp391]

Thank you all for help
But........
the new box is unchecked
Chart behavior the same as before
I am using cm93-2 from sept-2010
When I go back to 2.3.1 than the behavior is as usual

[bdbcat]

janp391/cagney....

OK, I see whats going on here. In certain parts of the cm93 database, the cells at a particular scale and location are split into multiple subcells. I think this was done in original cm93 to control file sizes.

We know how to load and display these subcells. However, the new code in 2.4 which caches cell M_COVR objects (and also manages user cell offsets) gets a bit confused by these subcells. They don't exist at the locations I used for testing, so I never saw this problem until we moved to more densely charted European locations.......

More code.....sigh....

Thanks for you efforts so far in identifying this problem.
Dave