OpenCPN Version 2.1.0 Beta

[sinbad7]

Hi all.
I have reverted back to ver. 1.3.6 and all my charts are functioning again.

It is either an enterpretation fault in the new version, OR perhaps OpenCPN now is more critical in it's acceptance of .kap charts,conforming more to the .bas/kap standard. Seems to be only Vista users who are affected.

However,I shall stay with v. 1.3.6 if this can not be resolved.

[GPS-Marco]

Hello.

I've seen that quilting of transverse mercator raster charts seems not working correctly. I attached the snapshot of the problem where a mercator chart has inside a TM chart. As you can see the TM chart is rendered too small. If the quilting is switched off both the mercator and the TM charts are rendered correctly so I guess the issue is related to rendering of TM charts in quilting mode.

Ciao, Marco.

[oem]

Originally Posted by oem

Here is another crash example. On my desktop vista machine the program starts, shows the initial 'ownship'-position and map, - but crashes after some 15-20 seconds without touching anything.

<cut>

Got it! It's the 'Course up' mark that causes crashes. As soon as I removed the 'Course up'-mark my vista desktop runs perfect, with or without my gps.

And whenever I mark the 'course up' it crashes again.

[bdbcat]

Folks....

OK, the following seem to be the critical bugs:

1. Course-up mode on Vista(Sinbad7, please test/confirm operation with course-up box de-selected on Worldmercator.kap). I'll try to reproduce this bug locally.

2. TMerc quilting: Acknowledged. Have spent zero time researching this issue. Priority is increasing.....

3. cm93 quilting: In design/coding now, coming soon I hope.

We may expect a new Beta within a few days, addressing at least some of the open bugs.

I remind you that if you need a stable opencpn platform for daily use, I recommend that you stay with 1.3.6. We are on the "Bleeding edge" with 2.1.0

Thanks for your reports and patience.
Dave

[nohal]

Originally Posted by bdbcat

1. Course-up mode on Vista(Sinbad7, please test/confirm operation with course-up box de-selected on Worldmercator.kap). I'll try to reproduce this bug locally.
Dave

I downloaded sinbad7's Worldmercator.cap to the w2k8 machine and with both course up on and off, besides the distortion seen on the screenshot, I experience a reproducible crash when zooming out with the following log messages

22:47:13: Initializing Chart C:\charts\Special\World\Worldmercator.kap
22:47:14:    Line Index corrupt on chart C:\charts\Special\World\Worldmercator.kap
22:47:14:    Assuming chart data is otherwise OK.
22:47:14:    VP Final Check: Georeference Chart_Error_Factor on chart C:\charts\Special\World\Worldmercator.kap is 0.0408965
22:47:26: x_malloc...malloc fails with request of -3565296 bytes.

Can't test on the XP machine right now as it's not traveling with me...

Pavel

[cagney]

Originally Posted by nohal

I downloaded sinbad7's Worldmercator.cap to the w2k8 machine and with both course up on and off, besides the distortion seen on the screenshot, I experience a reproducible crash when zooming out with the following log messages

22:47:13: Initializing Chart C:\charts\Special\World\Worldmercator.kap
22:47:14:    Line Index corrupt on chart C:\charts\Special\World\Worldmercator.kap
22:47:14:    Assuming chart data is otherwise OK.
22:47:14:    VP Final Check: Georeference Chart_Error_Factor on chart C:\charts\Special\World\Worldmercator.kap is 0.0408965
22:47:26: x_malloc...malloc fails with request of -3565296 bytes.

Can't test on the XP machine right now as it's not traveling with me...

Pavel

The problem with Sinbads world map is due to an incorrect kap file.
Running "bsbfix" on the file gives trhe following output

Warning: depth from IFM tag (6) != depth from bitstream (7)

Changing IFM from 6 to 7 fixed the garbled output, and correcting the KNP line, it was way over 80 chars long, got rid of the crash/malloc problem.

It is still badly georeferenced, with the red warning bar, but that is another problem.
I have attached the corrected version.

Thomas

[sinbad7]

Dave/Cagney..

I can not believe ALL my converted .kap charts are faulty. They all work perfectly with ver. 1.3.6 which I have reverted to for the moment.

Perhaps the new version has changed it's requirements of the .kap definition,and perhaps the older version was more 'forgiving' ?

If you have 1.3.6 installed and using Vista,try it out for yourself...

Can not imagine I have to alter ALL my .kap converted charts,in that case I'd rather remain with ver. 1.3.6

[sinbad7]

Dave/Cagney..

Changing IFM from 6 to value 7 made no difference at all for me.
In the WorldMercator.kap header,KNP is shown like this KNP/ with no following spaces.
The chart is only ment as an overview and I am aware of the calibration problem but then you will hardly navigate from it.

There must be another reason for the charts malfuctions,something is changed in the latest OpenCPN version.

I tested out your version of Worldmercator.kap and it loaded without distortion but in a strange scale. Trying to expand to a world view crashed the program. So it is NOT working in Vista. Thanks for trying Thomas.

[sinbad7]

Dave..

Welcome back by the way.. good to see you at the helm again ..

OK, the following seem to be the critical bugs:

1. Course-up mode on Vista(Sinbad7, please test/confirm operation with course-up box de-selected on Worldmercator.kap). I'll try to reproduce this bug locally.

Course up box,whether on or off makes no difference to my converted .kap charts.
There must be a difference between how v. 1.3.6 reads/displays .kap maps and the new version. However,it only concerns converted ,kap charts,the official .bsb charts displays normally.

[bdbcat]

Sinbad7, and others....

WorldMercator.kap indeed has a corrupted line index and improper color bit depth setting.

1.3.6 made the necessary corrections on the fly. I took that code out of 2.1.0 to help identify "questionable" charts as they come up. I guess I'll put it back.....

Other news: COG Up mode on Windows is definitely broken. Researching now.

Cheers
Dave

[nohal]

Originally Posted by cagney

Changing IFM from 6 to 7 fixed the garbled output, and correcting the KNP line, it was way over 80 chars long, got rid of the crash/malloc problem.

Garbling gone, the malloc problem stil there... Two ways to reproduce:
* Zoom out from TrueScale 38145200 Zoom 0.39
* Move the map (mouse or keyboard left/right couple of times) at TrueScale 38145200 Zoom 0.39

Originally Posted by bdbcat

1.3.6 made the necessary corrections on the fly. I took that code out of 2.1.0 to help identify "questionable" charts as they come up. I guess I'll put it back.....

Putting the correction routine back is a good idea as people don't like crashes even if they are part of the problem themselves

Pavel

[MarkJ]

"Course Up" closes my program.

With the GPS in it stops the GPS after 30 seconds and freezes.

[sinbad7]

Dave..

All my conversions have been done using the suggested available tools and have worked fine in previous versions. Perhaps you'd better put the old code back until corrections have been made to the existing conversion procedures.

I have been using 'Chartconvert' which uses mc2bsbh-tif2bsb-bsb2png-ImageMagick, after first converting the file to .png 127 col. in GIMP2

[sinbad7]

Dave..

For some strange reason,a few of my .kap conversions are displaying normally.
They were screendumps of C-Map with manual calibration from selected triangular points and converted with MapCal/Chartconvert. I wonder what is different with these charts?

Please download sample here:

[sinbad7]

Dave et all...

I am now running v. 2.1.0 and have just converted another .kap chart over
New York and approaches which is running perfectly?? I am using the same
method as always. Curiouser and curiouser...said Alice..