OpenCPN Version 1.3.5 Beta Technical

[sredna]

I noticed some issues related to marks and GPX usage:

After loading a gpx file containing a track, and then deleting the track, openCPN decided to keep the trackpoints as marks. Because they use the icon "Flag" (set by my garmin gps), which is unknown to openCPN, they can not be deleted from within the program. Consequently, I looked in ~/.opencpn, and deleted (after closing the app) a file named "navobj.xml", a gpx file containing the marks and routes in the program. But that did not help, so I looked in the config file, which contained an additional copy of those data in INI style.

Maybe storing only one place would be nice?

I also think that all marks that are not part of a route or track should have a visible representation.

Maybe it would be good to have a (larger) collection of common gps icons? There is a list on the gpsbabel.org site, and several icon collections around the web (though I do not know if they are legal to use).

A small issue with GPX loading/saving: When saving a route, the name entered in the route properties dialog is nicely saved in the route name in the gpx file, but it is not restored when loading.

Routes in exported GPX files have a bad element "prop", which is not in the GPX 1.1 schema, claimed to be used in the file. This means that the files are not valid, so should be fixed to ensure they are usable in other devices. The GPX schema has an "extensions" element which allows for such data.

[bdbcat]

BernardofCA:

Can you give me a lat/lon where this problem occurs? I've tried several places on linux and XP, no problem.

Do you mean that the detail control slider cannot be moved, or that it has no effect if moved?

Thanks
Dave

[bdbcat]

sredna....

Thanks for the feedback.

All good points. GPX development has been at low priority lately due to other pressing issues; not much utilization, I guess, since there has been little feedback. I suppose we should raise the priority a bit.

Regarding leftover waypoints after Track delete: there is another thread on the forum discussing this issue. Seems that some GPX files contain duplicated "trackpoints" and "waypoints". We don't feel comfortable deleting the "waypoints" along with the track, since we don't know if the waypoints belong exclusively to the track or not. Is this the case with your GPX files? Ideas on this issue are most welcome. How do other ECS's do this? What should we do?

Thanks, and happy new year....
Dave

[BernardOfCa]

Originally Posted by bdbcat

BernardofCA:

Can you give me a lat/lon where this problem occurs? I've tried several places on linux and XP, no problem.

Do you mean that the detail control slider cannot be moved, or that it has no effect if moved?

Thanks
Dave

Dave,

Actually I need to correct the description of the problem. It seems to have nothing to do with the OverZoom status.

The problem is the following: if the value in cm93 Detail is changed without dragging the cursor (by clicking anywhere on the slider bar), then the new value is not taken into account. This is different from the standard Windows behavior (for example volume cursors).

Is there a good reason for OverZoom to appear when the level of detail is changed?

I am playing in the LA turning basin at 33N 45.841 and 118W 15.247 with TrueScale 4000. OverZoom shows up at level < -3

Bernard

[bdbcat]

Bernard....

Good catch, thanks. You are correct. Keyboard control of the slider is not properly recognized. Will be fixed in Release coming.

Overzoom: If you lower the detail (move left), then you are requesting and moving to a smaller scale chart cell. Without changing the screen zoom, you then may indeed be overzoomed on the smaller scale cell. The logic triggers are the same as if you were zooming in on a chart for which there was no larger scale chart available, say in the middle of an ocean for instance.

Conversely, if you increase the slider, you may be "underzoomed" Nothing wrong with that per se, except that it takes longer to fetch and draw the chart, and there may be more detail (and clutter) than you really need. That's why we have a user controlled slider, instead of permanently setting the detail value to +5.

Thanks for the correction, and Happy New Year
Dave

[sredna]

Originally Posted by bdbcat

sredna....
All good points. GPX development has been at low priority lately due to other pressing issues; not much utilization, I guess, since there has been little feedback. I suppose we should raise the priority a bit.

Regarding leftover waypoints after Track delete: there is another thread on the forum discussing this issue. Seems that some GPX files contain duplicated "trackpoints" and "waypoints". We don't feel comfortable deleting the "waypoints" along with the track, since we don't know if the waypoints belong exclusively to the track or not. Is this the case with your GPX files? Ideas on this issue are most welcome. How do other ECS's do this? What should we do?

Right, I forgot that the file contained extra waypoints. So that part is solved...

I think using GPX/routes/waypoints has a lot of use cases, so I'm all for improvements in this department. Maybe I can help out a bit, at some point.

[blubaju]

Originally Posted by bdbcat

... Keyboard control of the slider is not properly recognized. ....

win7: I use 2 monitors (extended desk) and the frame usually pops up on the other monitor. Is there a possibility to tie this pop-up frame into the area of the main opencpn window? at least when it opens first time? user can still push it around afterwards

second question:, what about making it a little transparent, so not all background information is lost, might a new AIS target be hidden

Prosit Neujahr - Happy & Prosperous New Year!

[sinbad7]

ShipPlotter includes a feature whereby bearings can be displayed in
> degrees Magnetic or in degrees True. To convert between the two,
> ShipPlotter uses the World Magnetic Model (WMM) from the US National
> Geospatial-Intelligence Agency (NGIA).
>
> The coefficients for the model are gradually changing and are published
> by NGIA every five years. The last model (valid from 2005) expired at
> midnight on 31 December 2009. The new model coefficients are contained
> in the "WMM.COF" file which you will find in the files area of this
> group. Download the file and save it in the ShipPlotter application
> directory (typically C:\Program files\COAA\ShipPlotter\ but may be
> elsewhere on Vista/Win7 platforms) and restart ShipPlotter. The error
> message will then disappear for the next five years.

Dave.. Is OpenCPN in any way affected by this NGIA notification?

[Zoef]

Originally Posted by sinbad7

Pete..

Thanks for the tip. I downloaded the utility but I am very hesitant using it,as it MAY require re-installation of many or ALL programs previously installed using the Microsoft Installer.

Sinbad,
I have used this utility for several years, never had to re-install any program not selected for removal.
Klaas

[sinbad7]

Ref my # 413

If it applies to OpenCPN,here is the update file covering the period 2010-2015.. (Remove the .doc appendage)

[sredna]

Hi,
I am experimenting, and just now openCPN went bonkers with the lighthouse markings (sorry, I do not know the proper english word) as seen in the shot below. That is a good old fashioned bug :-)

[sredna]

... restarting openCPN made it go away. But still...

In more positive thoughts, what about replacing the vector map filter acronyms with the (translatable) human readable strings? I think I can manage to provide a patch for that.

[Ptizef]

Hi Dave

I found some small problems with "Depth Settings" , CM93 and Vista

1) Parameter "shallow Depth" set the medium_blue zone
If I choice 0 meter the zone disappear ----> OK
If I choice 1 meter the zone limit jumps to 5 meters soundings line
no difference between 1,...,5 : always 5 meters line
If I choice 6 meters the zone limit jumps to 10 meters line
no difference between 6,...,10 : always 10 meters line
so only 5 ,10 .. are possible
2) Parameter "Deep Depth" set the light_blue zone
the same behavior :
If I choice 10 meters zone limit OK at 10 meters line
If I choice 11 meters , the zone limit jumps to 20 meters line
no difference between 11,12....19,20 always 20 meters line
so only 10,20 ... are possible

3) Parameter "safety depth" set the zone where the soundings are displayed in strong black
when I set this parameter , it works well , but when opening opencpn , the value is not applied ,I have to open an close the tool box .
It seems that the parameter is not read when opening opencpn and a standard value of 5 is taken

Regards

Jean Pierre

[bdbcat]

Ptizef....
Regarding your points on "Depth Settings":

Points 1 and 2: The cm93 database contains depth area features according to fixed depth cutoff zones, usually 5, 10, and 20 meters. Intermediate values are not available in the database.
So, if you select a value between those available, opencpn chooses the next higher value available for display of color. This is by design.

Point 3: I cannot reproduce this. For me, the Safety Depth is properly stored in and recovered from the config file as:

S52_MAR_SAFETY_CONTOUR=4( or whatever...)

Anyone else having trouble setting/saving/restoring this parameter?

Dave

[bdbcat]

Hello again....

Here is opencpn Version 1.3.6 RTM

Changelog:
87. Change symbol for "UWTROC always dry" to same as "LNDARE as point feature".
88. Correct database logic preventing multiple duplicate chart additions.
89. Correct keyboard control logic for manipulation of cm93 detail slider.
90. Correct GPX output format using xmlnspencpn extensions.
91. Arrange to start opencpn with last displayed chart, scale, and zoom.
92. Update Help file
----Release 1.3.6----


New Link
http://www.southbaynetwork.com/openc..._136_setup.exe

CVS updated.

Notes:
Changes are small and localized, except....
90. Unavoidably large internal rework of GPX code to support maintenance and to ensure valid GPX I/O format. GPX users please test as fully as possible.

I have switched back to original Windows install .exe method. Reason: Vista will not install the 1.3.6 MSI version unless any current version of opencpn is removed first. Unfortunately, the 1.3.4 un-installer also removes and deletes the user data files, including config, SENCs, and all routes/tracks/waypoints. Not good. This fault has been corrected in the 1.3.6 un-installer, so we are free to move to the MSI method on the next production release if we desire.

I am posting this Release here first, as a Beta. This gives one last chance for correcting mortal flaws before I upload it intact to sourceforge, and as a preview for our loyal testers.

Going once, going twice.....

Thanks for your help
Dave