OpenCPN Beta Version 3.3.2316 Released

[NahanniV]

Uninstalled.
Deleted ProgramFiles(x86)\OpenCPN
ProgramData\OpenCPN was already gone

Re-installed.
Added single chart 12300

working OK.

Turn on OpenGL
Turn on Texture Compression with caching
Ship was on the chart.
Start to have delays when I zoom out to the base map.
Then get "waiting..." message on exit.

On restarting I tried Installing charts for all of NY state and Forcing the cache rebuild and it crashed almost immediately; I sent the crash-dump.

Cheers,
JM.

[seandepagnier]

Originally Posted by NahanniV

Uninstalled.
Deleted ProgramFiles(x86)\OpenCPN
ProgramData\OpenCPN was already gone

Re-installed.
Added single chart 12300

working OK.

Turn on OpenGL
Turn on Texture Compression with caching
Ship was on the chart.
Start to have delays when I zoom out to the base map.
Then get "waiting..." message on exit.

On restarting I tried Installing charts for all of NY state and Forcing the cache rebuild and it crashed almost immediately; I sent the crash-dump.

Cheers,
JM.

Are these noaa charts? What filenames? If you restart opencpn, and rebuild the cache again does it succeed and eventually work? I believe there are some race conditions with the threads building the cache.

[bdbcat]

jm...

How long doe the "waiting..." message on exit last? Does it eventually complete?

Dave

[NahanniV]

Q1/ Are these noaa charts?
A1/Yes the free NOAA charts

Q2/What filenames?
A2/initially I used all of NY State: http://www.charts.noaa.gov/RNCs/NY_RNCs.zip
After I tried with a single chart: 12300_1.KAP and 12300.BSB

Q3/If you restart opencpn, and rebuild the cache again does it succeed and eventually work?
A3/YES When I ran the compression again it seemed to work.

Q4/I believe there are some race conditions with the threads building the cache.
A4/There is Some problem.

[NahanniV]

Originally Posted by bdbcat

jm...

How long doe the "waiting..." message on exit last? Does it eventually complete?

Dave

Yes it completes in a few seconds.
It seems to be related to a general slowdown of the program.
The OpenCPN memory usage goes way up as well.

If the texture cache has been completely built, I don't get that message.

Thanks,
JM.

[RhythmDoctor]

Originally Posted by NahanniV

...Q2/What filenames?
A2/initially I used all of NY State: http://www.charts.noaa.gov/RNCs/NY_RNCs.zip
After I tried with a single chart: 12300_1.KAP and 12300.BSB

Sorry to ask a dumb question, but when you did all of NY State, did you unzip the file first into an uncompressed folder?

[NahanniV]

Originally Posted by RhythmDoctor

Sorry to ask a dumb question, but when you did all of NY State, did you unzip the file first into an uncompressed folder?

Ha Ha !

Yes, I did extract them from the Zip file.

Zip files are not even recognized by OpenCPN.

Cheers,
JM.

[bdbcat]

JM...

We changed the naming scheme for the raster texture cache. So Beta testers will notice reduced performance and higher memory usage until their cache is rebuilt with the new names.

There may still be trouble with the bulk cache build option. I will look at your crash report and see what's up.

Thanks
Dave

[NahanniV]

Originally Posted by bdbcat

JM...

We changed the naming scheme for the raster texture cache. So Beta testers will notice reduced performance and higher memory usage until their cache is rebuilt with the new names.

There may still be trouble with the bulk cache build option. I will look at your crash report and see what's up.

Thanks
Dave

But, I am also seeing a problem with only 1 chart installed. Once the memory usage goes up and the system slows down it continues till exiting OpenCPN and then I get the "Waiting on Compression Thread Exit" message.

The bulk Caching crash may have happened when the the program was already in the slow state.

Cheers,
JM.

[bdbcat]

JM...

What is your system?

On Windows, OCPN will use up to 1 GB of physical memory, usually, unless the total available memory is 1GB or less. Then he uses up to 50% of available physical system memory.

OCPN does not attempt to reduce its footprint below the limits established above. More memory for OCPN means better performance, so we use all we can (reasonably).

Dave

[NahanniV]

Originally Posted by bdbcat

JM...

What is your system?

On Windows, OCPN will use up to 1 GB of physical memory, usually, unless the total available memory is 1GB or less. Then he uses up to 50% of available physical system memory.

OCPN does not attempt to reduce its footprint below the limits established above. More memory for OCPN means better performance, so we use all we can (reasonably).

Dave

Win 7 4GB

[bdbcat]

All...

There was some bug reported about showing invalid ownship position upon returning from sleep/hibernate mode. I assume on Windows, but details are sparse....

Anyone have any more info or scenario to reproduce this?

Thanks
Dave

[Hakan]

Crash report WinXP Beta version: 2014-11-28 last commit: “Merge pull request #202...…………….”

I've used both raster and CM93 –charts and the crash seems to come when "fogging" shall start.

FogOnOverzoom=1
Crash when zooming in. Last scale before crash: Scale 5100 (9.7x) (Seems to be a limit at 10x)
The crash occur independent of raster or vector charts usage.

FogOnOverzoom=0
Now I can zoom more then 10x and no crash.


OpenGl settings from my log-file:

  09:37:49: Warning: unable to use NPOT texture for chart symbols
  09:37:50: Fel: XPM: Trunkerat bilddata på rad 4!
  09:37:50: OpenGL-> Renderer String: MOBILITY RADEON X600 x86/SSE2
  09:37:50: OpenGL-> Texture rectangle format: 0
  09:37:50: OpenGL-> Using Vertexbuffer Objects
  09:37:50: OpenGL-> Framebuffer Objects unavailable
  09:37:50: OpenGL-> Using Stencil buffer clipping
  09:37:50: OpenGL-> Using s3tc dxt1 compression
  09:37:50: OpenGL-> Compressed tile size: 128kb (6:1)
  09:37:50: OpenGL-> Minimum cartographic line width:  1.0
  09:37:50: OpenGL-> Minimum symbol line width:  1.0
  09:37:50: ChartDB Cache policy:  Application target is 1023 MBytes

Håkan

[Hakan]

Originally Posted by Hakan

Crash report WinXP Beta version: 2014-11-28 last commit: “Merge pull request #202...…………….”

And the Crash is still there after today's pull including commit: "Add FullScreen toggle to....................."

So I would say; when "bool fog_it" is true there's a crash.
Håkan

[Hakan]

Originally Posted by Hakan

And the Crash is still there after today's pull including commit: "Add FullScreen toggle to....................."

So I would say; when "bool fog_it" is true there's a crash.
Håkan

And some debug break info:
First I recieve a break, which "destroyed my "Synaptics TouchPad" slider funcion, but this is only at Debug not in run time:

  opencpn.exe': Loaded 'C:\Program\wfind2k\WFHotMouse.dll', Binary was not built with debug information.
  opencpn.exe has triggered a breakpoint 

  
  Brake in “appbase.cpp” row: 678 
  675 #else
  676    // TODO
  677 #endif // Win/Unix
  678 } 

But I can continue and then when I zoom in far enough to make "Fog_it"=1 the following brake uccur:

  Unhandled exception at 0x692b65ec in opencpn.exe: 0xC0000005: Access violation reading location 0x00000000.
  Brake in OCPNregion.cpp: 2623
       /* some ATI drivers require this filter mode to build MipMaps, so to be safe... */
                      glTexParameteri( GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_LINEAR );

And from here I can't continue.
Håkan