OpenCPN Version 2.4 Beta Build 412

[bdbcat]

Ahoy OpenCPN'ers...

OpenCPN Version 2.4 Beta Build 412 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
31. Correct cm93 quilt rendering errors when mouse dragging and clicking.
32. Improve useability of cm93offset dialog.(FS#308)
33. Ignore wgsox/y embedded offsets in cm93 database, while preserving user offset capability.(FS#393)
34. Correct vector chart rendering of soundings drawn very near screen edge. (FS#???)
35. Correct calculation and rendering of Lat/Lon grid.(FS#391)
26. Correct logic for chart database update requirement. (FS#370)
27. Correct AIS ROTIND = +/-127 target query report. (FS#388)
28. Implement improved chart quilt-patch highlighting. (FS#395) Thanks Will.
29. Improve S57 ENC behavior in Full Screen Quilting mode. (FS#397) Thanks to Will for Youtube demo.
30. Implement "esc" key for route creation end in addition to context menu "End Route". Thanks HappySeagull
31. Integrate Enhanced Tide/Current Info patch. Thanks jp. (FS#128)
32. Integrate Automatic Daily Tracking patch (FS#???) Thanks piotr
33. Integrate Version 2 of RouteManagerDialog Layers Tab. Thanks, piotr.
34. Correct off screen object move logic. (FS#347)
35. Implement AIS ATON decoding and reporting. (FS#340) Thanks piotr
36. Improve implementation of Screen Brightness (F6) function.
37. Implement "Full Screen" toggle (F10)
38. Improve Quilt logic for large scale range quilt. (FS#316)
39. Implement persistent quilting algorithm. (FS#398) Thanks Will, for the idea.
40. Implement "confirm" dialog for track delete from context menu.
41. Implement enhanced RoutePlanner function, thanks to piotr, again....
42. Accept header parameter "PR=TM" for TMerc BSB charts, as well as "PR=UTM".
43. Implement AIS Inland binary broadcast, DAC 200 FI 10, especially UN Ship Type (FS#385).
44. Correct behavior of AIS target list rangle limit spin control (FS#355)
45. Correct AIS target list sorting limits. Sorting disabled with more than 1000 targets. (FS#339)
47. Correct rendering of some vector chart features, especially isolated small rocks and depth contours. (FS#373)
48. Save and Restore "Ships Bells" state.
49. Correct symbolization of Moored and Anchored AIS targets. (FS#389)
50. Improve chart info window rollover formatting (FS#249)
50. Change Grid legend color to light grey.
51. Various language file updates.
52. Modify plugins/CMakeLists.txt to streamline plugin building. Thanks to Jean-Eudes for the idea.

Release Beta Version 2.4.412


Notes:

31. This is neat. A big thanks to JP for the programming effort. Lots of fiddly stuff required here. Applause.....

33. I'm just coming up to speed on the Layers function. Integrated as received. Feedback appreciated.

36/37. Highly platform dependent. Probably not right yet for non-Windows platforms. Fussy....

40. Suggestion from down under. Good idea. Tracks are valuable and potentially irreplacable....

47. This one will be important. You will note the enforced symbolization of Safety Depth contour (or nearest greater contour, depending on the contours available in the ENC) in all Display Categories. Also, UWTROC is always symbolized if dangerous with respect to safety depth and attributes.

51. Languages: I updated what I have received recently. I am not making a great effort to keep .po and .pot files up to date until we get a little closer to release of Version 2.5.xxxx



Good Luck
Dave

[r.fairman]

Dear Dave

Thanks for the new beta Can you add a request to implement F1 help with context sensitive Help in the next beta. I can convert existing documentation to Help format The developers need to implement F1 and supply context ids for the various screen and dialog boxes It would be nice to include this in the next beta for testing. I know this is the wrong place to ask this but I do not know how to fromally ask for a feature

regards

Richard Fairman

[balp]

Originally Posted by bdbcat

37. Implement "Full Screen" toggle (F10)
...
36/37. Highly platform dependent. Probably not right yet for non-Windows platforms. Fussy....

Works fine in Linux Ubuntu. This funcitons I really need on my netbook, 800x480 is not much screen space.

[nohal]

37. Implement "Full Screen" toggle (F10)
...
36/37. Highly platform dependent. Probably not right yet for non-Windows platforms. Fussy...

Interesting...
Looking at the code (no wxFULLSCREEN_* in the style), it should work the way it does for me on Windows - just maximize the window without removing the toolbar, the border or the window caption.
For some reason it does work as expected on Linux - removes everything but the statusbar.
Made some effort to achieve the same on Win and ended up with the following code

void MyFrame::ToggleFullScreen()
{
      bool to = ! IsFullScreen();
      if (!to)
      {
            ShowFullScreen(to, wxFULLSCREEN_NOBORDER | wxFULLSCREEN_NOCAPTION | wxFULLSCREEN_NOTOOLBAR);
            m_toolBar = CreateAToolbar();
            SetToolBar((wxToolBar*)m_toolBar);
            Iconize();
            Restore();
            Raise();
      }
      else
      {
            DestroyMyToolbar();
            ShowFullScreen(to, wxFULLSCREEN_NOBORDER | wxFULLSCREEN_NOCAPTION | wxFULLSCREEN_NOTOOLBAR);
      }
}

Why destroying and recreating the toolbar? If it's not done, there is a toolbar artifact in the fullscreen mode. Don't know why.
Why that Iconize, Restore, Raise hack? After returning from fullscreen mode the toolbar was not drawn properly and the chart viewport was not resized. To be honest I have no idea why. What helps though is resizing the window - to keep the previous size, I used this "trick".

Tested on XP, 7 and Ubuntu 10.10/32-bit

Ugly is this stuff...

Pavel

[sinbad7]

Dave..

Is beta 412 released or what? Can't seem to find it on opencpn.org.

Tore

[cagney]

Ububtu versions of 2.4.412 are now available.
Thomas

[cagney]

Dave

Thank you very much for all the hard work, including the night watches, to launch another great beta.

Regarding full screen. I see different behavior in the two hardware setups I'm running.
Ubuntu 32 bit Atom on an Eeepc901. F10 works exactly as I like, only the chart and status bars are left, the toolbar goes. When switching "show statusbar" on/off I get a crash. Dump below.



Ubuntu 64 bit amd. First the toolbar goes, but comes back after a second. I can´t repeat the crash, described above.

Dimming with F6 doesn't work at all on either platform, but there are other methods available on Linux.

Thomas

[PjotrC]

Thanks for hard work and amazing richness of new features! I truly look forward to using this for planning and underway soon!

Windows Vista compiles OK, but when running, crashes too often (just moving and rezooming) to even attempt at reproducing... After many restarts seems to stabilize a bit, at least in my area of interest of Aland Archipelago (quite complex cartography there). This is exhibit 1, nothing really wrong there, but initially _much_ slower than 1.3.6...

2. I admit to have tried F10 lightheartedly, so now I do not recall the pristine behaviour, but the persistent extra black window, having some interaction with my wireless Internet connection app seems not quite OK...

3. the UWTROC (I assume) line of 50m depth (I understand this is not settable?) does not help me at all, I definitely prefer the 1.3.6 chart appearance

4. At some scale zoom there is a strange coastline contour, probably not needed there

Very exciting release indeed!

Thanks again so much,

Piotr

[mrm]

Source code documentation for the current beta generated and published here: OpenCPN: 2.4.412 source code documentation

[PjotrC]

Originally Posted by PjotrC

Windows Vista compiles OK, but when running, crashes too often (just moving and rezooming) to even attempt at reproducing...

The most sensitive part seems to be changing the Toolbox settings, especially CM93 Detail Level slider...

Excellent track etc. rendering speed... gives a new feeling!

[nohal]

Piotr,
don't know if it's related to your crashes, but it looks like we might have a memory leak, which gets worse in areas heavily "populated" with symbols (like Aaland)...
Not sure how to track it down to more than

Detected memory leaks!
Dumping objects ->
{840728} normal block at 0x06E77E78, 56 bytes long.
 Data: <    LIGHTS      > 00 00 00 00 4C 49 47 48 54 53 00 00 00 00 00 00 
{839835} normal block at 0x06E77B00, 56 bytes long.
 Data: <    SOUNDG      > 00 00 00 00 53 4F 55 4E 44 47 00 00 00 00 00 00 
{839711} normal block at 0x06E77788, 56 bytes long.
 Data: <    SOUNDG      > 00 00 00 00 53 4F 55 4E 44 47 00 00 00 00 00 00 
{839641} normal block at 0x06E77410, 56 bytes long.
 Data: <    SOUNDG      > 00 00 00 00 53 4F 55 4E 44 47 00 00 00 00 00 00 
{839463} normal block at 0x06E770C0, 56 bytes long.
 Data: <    SOUNDG      > 00 00 00 00 53 4F 55 4E 44 47 00 00 00 00 00 00

It goes further depending on how much stuff there is on the charts displayed.
Valgrind on Linux also reported a definite leak in s52plib.cpp(6551) but this part of the code is something I really don't understand at all...
When tried to uncomment the pAlloc->Add ( NewLUP ); on the next line, it looks fixed.

Pavel

[Ptizef]

Tested Pavel solution (#4 above) on Vista and I get a crash each time I use a function key (F2 for example) when the tool bar is hidden
The best result is optained with this code :

void MyFrame::ToggleFullScreen()
{
bool to = ! IsFullScreen();
//ShowFullScreen(to, 0);//wxFULLSCREEN_NOBORDER | wxFULLSCREEN_NOCAPTION);

if (!to)
ShowFullScreen(to, 0);
else
ShowFullScreen(to, wxFULLSCREEN_NOBORDER | wxFULLSCREEN_NOCAPTION );
}

but I have still that odd issue with F10 key ( see FS#403)
Jean Pierre

[nohal]

Right you are
Of course all the function keys that try to do something with the toolbar (which I destroy previously) must crash the program... Easy fix.
How does Vista like the attached patch (done against vanilla git)?

Pavel

[Ptizef]

Hi Pavel
Tested your patch . Works fine
but is it the good solution to hide tool bar in full screen mode ?
From my point of view , I'd prefer to keep it displayed .
Or the best but the most complex solution would be to display it only when the cursor rolls over the top part of the screen (as most of media readers) ?

Am I the only one to have the F10 key deactivating other function keys ?

regards
Jean Pierre

[nohal]

What I tried to do with this patch is to have the same behavior on Windows like I have on Linux, which means no toolbar at all.
The other side of things is, that while it was easy to do this, I can't make the bloody thing actually do what you suggest (which of course was my original plan) But I think I'm very slowly starting to understand what happens here as I had very similar problems while working on the GE plugin.
My F-keys look working quite OK - there is no reason why they shouldn't looking at the code, so let's wait if it appears for more people.

Pavel