OpenCPN Beta Version 3.1.802 Released

[P_Dub]

First I lost my file, that I should have prevent this; but, look at the diiferences in the interface, you will see yourself the changes. I was getting ready to send a correct tidal predictions and not what you hav at the moment. Too many differences with the official tides and current as stated on the charts. For example, my setup was with the tidal differences on Dover High Water for the English Channel, which is always what was used by seaman. I had cover from the Scilly Isles to Sandetti Bank. I had only a few stations left to go to Germany. What a shame it is.
Bye.

[bdbcat]

P-Dub...

Please calm down, start over, and tell us what you have been working on, and how 3.1.802 broke that.

Thanks
Dave

[P_Dub]

I succeded to rcuperate part of it only, I am not blaming you for it.
Line 20 scaling of AIS Target does not work and also the panning and zooming is not capable of being check ( its grey ) in the Windows version.
Very slow upon moving to the next chart (with 8 gigs). It needs improving for the memory usage.
I was almost getting ready to send you the corrct files for tide predictions for European Waters, as the ones you have are very doubtfull and not correct time and height for many stations. England for example is not working on 2 hours on GMT.
Bye.

[JesperWe]

P_Dub, I can only speak for myself but I think we all have a bit of trouble understanding what you are referring to. Please try to be a bit more exact and try not to mix comments about multiple issues in the same sentence.

What is the "it" that you could only recuperate parts of?

For "panning and zooming", do you mean the "Enable smooth panning/zooming" check-box? Maybe you have the "Enable Wheel Zoom-to-cursor" check-box checked? Then you cannot enable smooth panning/zooming. This has always been the case, in previous versions too.

Describe the charts you have loaded (raster/vector, how many, sizes etc) and which ones are displayed in your "Very slow upon moving to the next chart" situation. "Moving" to me means panning, is that correct? Or do you mean zooming or changing scale? Or maybe using the chart status bar buttons?

You see, you have to be more specific, otherwise all we know is that you are upset, and not really why...

[cagney]

It's very hard to see how any changes to the old tidal files could be lost.
One of the smart things with this update, is that the old format is kept parallel to the binary tcd format.
Files with the two different format can be loaded at the same time, and the expanded text field in the dialog states which file is responsible for the particular station. This helps in evaluating the tidal data.
There are clear advantages with the tcd format. A few of them:

*David Flater of Xtide regularly releases updated files, for the tidal stations with available free data. While Ubuntus tcd files for xtide is from 2010, there are files from 2011 and 2012 available at XTide Downloads. Note that this includes US, UK and the Netherlands.

* There is a graphical program "tidEditor" available for updating old stations and creating new ones. Same link as above. This is very much easier than tweaking the old formats huge text files manually.

*OpenCPN now supports the format (tcd) used by many other navigation programs.

A few tcd files are available at http://http://wave.oceancares.org/tides.html

Thomas

[JesperWe]

Dave, could the slowness P_Dub is referring to be due to the cash memory limit fix still not being applied?

diff --git a/src/chart1.cpp b/src/chart1.cpp
index 36d21d3..29d55c1 100644
--- a/src/chart1.cpp
+++ b/src/chart1.cpp
@@ -1375,7 +1375,7 @@
         //  not to exceed 1 GB
 #ifdef __WXMSW__
     if( 0 == g_memCacheLimit ) {
-        if( CACHE_N_LIMIT_DEFAULT == g_nCacheLimit ) g_memCacheLimit = wxMin((int) ( g_mem_total * 0.5 ), 1024);
+        if( CACHE_N_LIMIT_DEFAULT == g_nCacheLimit ) g_memCacheLimit = (int) ( g_mem_total * 0.5 );
     }
         g_memCacheLimit = wxMin(g_memCacheLimit, 1024 * 1024);  // math in kBytes
 #endif

[P_Dub]

I will redo the work, but it is taking times. The data I have are based on analysing the actual recorded data for each stations. By special permission from each Authority and I can tell you that there are many differnces. The gentleman you are refering to is probably knowing about it and does all he can to update it. But it require a fair amount of work to veryfied each stations in order to have future predictions fairly correct.
Regards.

[bdbcat]

Dear Windows testers...

There is a bug in 3.1.802 chart memory management module.

There is a workaround:
Place the following two lines in your opencpn.ini file.

[Settings]
NCacheLimit=19
MEMCacheLimit=512

This will force a target OCPN memory usage limit of 512 MBytes, and should be reasonable for testing on most machines.

The correct solution to this bug has been pushed to github branch master.

Thanks
Dave

[bdbcat]

Thomas, and others...

Concerning tide calculations using legacy .IDX files on 3.1.802:

I cannot reproduce any problem. It all works fine for me. I traced the code, and it all looks fine.

Here is a key point of evidence: The bad graphs you have shown each has no units indicated on the left side of the curve. Should be 'meters' or 'feet'. If you see no such indication, the calculation is incorrect.

Are your results absolutely repeatable? Does the same thing every time?

Dave

[fgd3]

I'm running OpenCPN 3.1.802 on Windows Vista and Windows 7. When I narrow the window beyond the point where the toolbar and the GPS indicator will both fit the GPS indicator gets left behind. It still appears on the screen but it doesn't move with the rest of the window. If I drag the window over the GPS indicator the GPS indicator disappears. If I make the window wider the GPS indicator reappears where it should be. If I have moved the window lower in the meantime the left-behind GPS indicator "jumps" to the window as soon as there is space for it. Has anyone else noticed that?

Fabbian

[sailorF54]

Originally Posted by bdbcat

Thomas, and others...

Concerning tide calculations using legacy .IDX files on 3.1.802:

I cannot reproduce any problem. It all works fine for me. I traced the code, and it all looks fine.

Here is a key point of evidence: The bad graphs you have shown each has no units indicated on the left side of the curve. Should be 'meters' or 'feet'. If you see no such indication, the calculation is incorrect.

Are your results absolutely repeatable? Does the same thing every time?

Dave

Here is the .idx files many of us are using in France, and that give meaningless predictions on 3.1.802

[JesperWe]

Originally Posted by fgd3

I'm running OpenCPN 3.1.802 on Windows Vista and Windows 7. When I narrow the window beyond the point where the toolbar and the GPS indicator will both fit the GPS indicator gets left behind.

Yup I see that. That's a regression from last beta...

[kubek]

Hi,

I try to play with Route Properties, everything works fine (Color, Width) but "Style" doesn't work at all (see attachment).

Pawel

[nohal]

Pawel...
The line styles just work in non-OpenGL mode.

Pavel

[P_Dub]

I decided to revert to 3.1.714 for the moment until you sort out the problem with using previous style harmonic.All is working fine with that version.
Regards and good luck.