Beta Test / Technical

[PjotrC]

sinbad7...

Do you have samples of the data feed from various points in the process of transmission / conversion / storage? Does this issue concern only OpenCPN? If so, I would look at it with properly instrumented code, if I can have some data.

[sinbad7]

PjotrC..

Unfortunately I only have one major data feed from the Siitech network so I am unable to produce separate outputs for the ports mentioned earlier. Yes, I have the same problem using other nav progs.

If I use AISHUB data feed,all the ship's data becomes available after a few minutes.

[Ptizef]

Hi
I've compiled route manager on window vista
But the problem : after opening route manager dialog , there is no route in it and if I click 'import' , and choose a route in the list , the route is imported but the dialog stay empty . May be I've misunderstood something ?
Regards
Jean Pierre

[Ptizef]

Furthemore , when routes have been imported or created by the official opencpn version , they are not visible in the patched version , but the contrary is good
Jean Pierre

[idpnd]

This might be a windows-related bug with the patch, the routes worked great for me on ubuntu

[Psyches]

Originally Posted by idpnd

This might be a windows-related bug with the patch, the routes worked great for me on ubuntu

Thanks idpnd, that's my cue to take a look. Jean Pierre & all, I see the same problem on win32.

Anders, listctrl is getting an assert during debugging. Here is the stack trace:

 	ntdll.dll!_DbgBreakPoint@0() 	
>	opencpn.exe!wxTrap()  Line 678	C++
 	opencpn.exe!DoShowAssertDialog(const wxString & msg={...})  Line 753	C++
 	opencpn.exe!wxAppTraitsBase::ShowAssertDialog(const wxString & msgOriginal={...})  Line 557 + 0x9 bytes	C++
 	opencpn.exe!wxGUIAppTraitsBase::ShowAssertDialog(const wxString & msg={...})  Line 635	C++
 	opencpn.exe!ShowAssertDialog(const char * szFile=0x00bf0a00, int nLine=0x0000065d, const char * szFunc=0x077f6ba4, const char * szCond=0x00bf09bc, const char * szMsg=0x00bf0990, wxAppTraits * traits=0x01c5a330)  Line 836 + 0x11 bytes	C++
 	opencpn.exe!wxAppConsole::OnAssertFailure(const char * file=0x00bf0a00, int line=0x0000065d, const char * func=0x077f6ba4, const char * cond=0x00bf09bc, const char * msg=0x00bf0990)  Line 445 + 0x22 bytes	C++
 	opencpn.exe!wxOnAssert(const char * szFile=0x00bf0a00, int nLine=0x0000065d, const char * szFunc=0x00bf09e8, const char * szCond=0x00bf09bc, const char * szMsg=0x00bf0990)  Line 713	C++
 	opencpn.exe!wxListCtrl::InsertItem(const wxListItem & info={...})  Line 1629 + 0x43 bytes	C++
 	opencpn.exe!RouteManagerDialog::UpdateRouteListCtrl()  Line 239	C++
 	opencpn.exe!RouteManagerDialog::RouteManagerDialog(wxWindow * parent=0x027080a8)  Line 202	C++
 	opencpn.exe!ChartCanvas::ShowRouteManager()  Line 1175 + 0xf bytes	C++
 	opencpn.exe!ChartCanvas::PopupMenuHandler(wxCommandEvent & event={...})  Line 4095	C++
 	opencpn.exe!wxAppConsole::HandleEvent(wxEvtHandler * handler=0x027080a8, void (wxEvent &)* func=0x00677fea, wxEvent & event={...})  Line 323	C++
 	opencpn.exe!wxEvtHandler::ProcessEventIfMatches(const wxEventTableEntryBase & entry={...}, wxEvtHandler * handler=0x027080a8, wxEvent & event={...})  Line 1233	C++
 	opencpn.exe!wxEventHashTable::HandleEvent(wxEvent & event={...}, wxEvtHandler * self=0x027080a8)  Line 907 + 0x1c bytes	C++
 	opencpn.exe!wxEvtHandler::ProcessEvent(wxEvent & event={...})  Line 1293 + 0x1c bytes	C++
 	opencpn.exe!wxMenuBase::SendEvent(int id=0x00000008, int checked=0x00000000)  Line 792 + 0x1c bytes	C++
 	opencpn.exe!wxMenu::MSWCommand(unsigned int __formal=0x00000000, unsigned short id=0x0008)  Line 792	C++
 	opencpn.exe!wxWindow::HandleCommand(unsigned short id=0x0008, unsigned short cmd=0x0000, void * control=0x00000000)  Line 4970 + 0x12 bytes	C++
 	opencpn.exe!wxWindow::MSWWindowProc(unsigned int message=0x00000111, unsigned int wParam=0x00000008, long lParam=0x00000000)  Line 2924 + 0x22 bytes	C++
 	opencpn.exe!wxWndProc(HWND__ * hWnd=0x00010a76, unsigned int message=0x00000111, unsigned int wParam=0x00000008, long lParam=0x00000000)  Line 2618 + 0x1c bytes	C++
 	user32.dll!_InternalCallWinProc@20()  + 0x28 bytes	
 	user32.dll!_UserCallWinProcCheckWow@32()  + 0xb7 bytes	
 	user32.dll!_DispatchMessageWorker@8()  + 0xdc bytes	
 	user32.dll!_DispatchMessageA@4()  + 0xf bytes	
 	opencpn.exe!wxYieldForCommandsOnly()  Line 2175	C++
 	opencpn.exe!wxWindow::DoPopupMenu(wxMenu * menu=0x072a1b58, int x=0x00000361, int y=0x000001eb)  Line 2224	C++
 	opencpn.exe!wxWindowBase::PopupMenu(wxMenu * menu=0x072a1b58, int x=0x00000361, int y=0x000001eb)  Line 926 + 0x41 bytes	C++
 	opencpn.exe!ChartCanvas::CanvasPopupMenu(int x=0x00000361, int y=0x000001eb, int seltype=0x00000005)  Line 4003	C++
 	opencpn.exe!ChartCanvas::MouseEvent(wxMouseEvent & event={...})  Line 3812	C++
 	opencpn.exe!wxAppConsole::HandleEvent(wxEvtHandler * handler=0x027080a8, void (wxEvent &)* func=0x0067dc38, wxEvent & event={...})  Line 323	C++
 	opencpn.exe!wxEvtHandler::ProcessEventIfMatches(const wxEventTableEntryBase & entry={...}, wxEvtHandler * handler=0x027080a8, wxEvent & event={...})  Line 1233	C++
 	opencpn.exe!wxEventHashTable::HandleEvent(wxEvent & event={...}, wxEvtHandler * self=0x027080a8)  Line 907 + 0x1c bytes	C++
 	opencpn.exe!wxEvtHandler::ProcessEvent(wxEvent & event={...})  Line 1293 + 0x1c bytes	C++
 	opencpn.exe!wxWindow::HandleMouseEvent(unsigned int msg=0x00000204, int x=0x00000361, int y=0x000001eb, unsigned int flags=0x00000002)  Line 5157 + 0x1c bytes	C++
 	opencpn.exe!wxWindow::MSWWindowProc(unsigned int message=0x00000204, unsigned int wParam=0x00000002, long lParam=0x01eb0361)  Line 2884 + 0x21 bytes	C++
 	opencpn.exe!wxWndProc(HWND__ * hWnd=0x00010a76, unsigned int message=0x00000204, unsigned int wParam=0x00000002, long lParam=0x01eb0361)  Line 2618 + 0x1c bytes	C++
 	user32.dll!_InternalCallWinProc@20()  + 0x28 bytes	
 	user32.dll!_UserCallWinProcCheckWow@32()  + 0xb7 bytes	
 	user32.dll!_DispatchMessageWorker@8()  + 0xdc bytes	
 	user32.dll!_DispatchMessageA@4()  + 0xf bytes	
 	opencpn.exe!wxEventLoop::ProcessMessage(tagMSG * msg=0x0012fc4c)  Line 80	C++
 	opencpn.exe!wxEventLoop::Dispatch()  Line 294	C++
 	opencpn.exe!wxEventLoopManual::Run()  Line 115 + 0xd bytes	C++
 	opencpn.exe!wxAppBase::MainLoop()  Line 312 + 0x15 bytes	C++
 	opencpn.exe!wxAppBase::OnRun()  Line 368	C++
 	opencpn.exe!wxEntryReal(int & argc=0x00000002, char * * argv=0x01c57b18)  Line 460 + 0x1b bytes	C++
 	opencpn.exe!wxEntry(int & argc=0x00000002, char * * argv=0x01c57b18)  Line 209 + 0xd bytes	C++
 	opencpn.exe!wxEntry(HINSTANCE__ * hInstance=0x00400000, HINSTANCE__ * __formal=0x00000000, HINSTANCE__ * __formal=0x00000000, int nCmdShow=0x00000001)  Line 386 + 0xd bytes	C++
 	opencpn.exe!WinMain(HINSTANCE__ * hInstance=0x00400000, HINSTANCE__ * hPrevInstance=0x00000000, char * lpCmdLine=0x00151f0d, int nCmdShow=0x00000001)  Line 736 + 0x33 bytes	C++
 	opencpn.exe!__tmainCRTStartup()  Line 578 + 0x35 bytes	C
 	opencpn.exe!WinMainCRTStartup()  Line 403	C
 	kernel32.dll!_BaseProcessStart@4()  + 0x23 bytes

I'll need to take a closer look when I get some time, if someone else doesn't first. BTW, when the route manager becomes ready for prime-time, I suggest we co-opt the "create route" toolbar button, and add a "New" button to the route manager, and/or add "Create Route" to the context menu. Interested in opinions.

Mark

[idpnd]

I suggest we co-opt the "create route" toolbar button, and add a "New" button to the route manager,and/or add "Create Route" to the context menu. Interested in opinions.

Clever!

[Psyches]

Hi Anders, haven't heard from you in awhile. I finally got a few minutes to look at the route manager, and why it fails on win32. I haven't had time to make everything work, only a few basics. And, I have a few suggestions. A few of the problems I found:

(1) for initial run, the IsVisible should default to true - otherwise initially existing routes don't show, which I think is confusing for folks
(2) image display didn't seem to work for some reason
(3) every item insert into the listctrl fails (the assert I mentioned)
(4) the column sizes aren't right sometimes - e.g. if no column #2 data, the header is messed up
(5) route selection didn't work at all
(6) the dialog sizer combination doesn't properly position the ok button initially - it partially overlays another button
(7) when I got route selection working, the button state update didn't work

My overall suggestion is: you have really good ideas and good implementation, but I suggest you keep things simpler for first iterations, else you may be surprised at how difficult it is to get win32, unix, and mac all working.

(1) Initially it wasn't obvious if something basic was broken - turns out it's just that old routes start out invisible...

(2) you may need to provide two images, one for invisible, the other for visible...dunno for sure yet; that is, having no image for invisible, and an image for visible may be causing a problem?

(3) I'm no ListCtrl expert, but in the past I've found it's pretty finicky on win32. The SetColumn(1) call you made causes all inserts to fail...that was the assert issue. Instead you need to specifically set the name in col 1 as you do the route "description". That also means the SetText call can't be made, else the name shows up in the wrong column. There may be a better way around this - with a bit of experimenting or ListCtrl expert advice.

(4) The win32 column autosize doesn't take header into account, so if there's no data, the header looks strange (there, but zero length). You could either fix this by providing default text or conditionally setting the size to fixed.

(5) the hittest always fails for me in the mouse left down event. It would be much simpler to use the built-in event support for item selected, activated, etc. I was able to add an OnSelected method in a few minutes that worked the first time at calculating the selected item. I'm not sure how long it'd take to debug the mouse click + hittest approach - more time than I have right now.

Also, FWIW, although the special control handling is cool, I tend to dislike "secret key combos", and prefer dedicated buttons to do the work...that way it's obvious to folks w/o doc and a good memory. I think supporting both would be a fine approach, after the basics work.

And also FWIW, while I think your approach using ::Connect has merit, it also has the disadvantage that you have to scour the code to find which events are handled, and where. The advantage of the event table approach is that it's effectively declarative. It's faster for maintenance and code assessment to see what events are handled and where, up front. Plus that's the way the rest of OpenCPN does it...

I've attached code that no longer asserts and properly sees selections, but I still haven't worked through a few of the problems I ran into above.

Mark

[Ptizef]

Hi ,Mark
The routemanager-tweak works a little better that the first one :
now , when I import routes , they are visible in the dialog , but all except import function are still inactive (except accidentally sometime ?)
The function visible/invisible works except for the 1st line when all routes are invisible
When activate a route , there is no active wp blinking
The end legs arrows are curious : some are very big , others very small
regards
Jean Pierre

[PjotrC]

I think Mr. Mercator may have a problem with small values.
The original 1.3.6 too, but it is easier to see here.
Until anchor chain lengths came into the picture it was no issue...

Piotr

[HamishB]

Originally Posted by PjotrC

I think Mr. Mercator may have a problem with small values.
The original 1.3.6 too, but it is easier to see here.
Until anchor chain lengths came into the picture it was no issue...

could you further describe what we are looking at there and what is wrong? (I don't know the software well)
130m looks shorter in the x than in the y?

what latitude are you at, what's the map scale?

[Totobeloeil]

You'll have to go South for better results ...
I don't know if it's interesting to consider only the distance problem.
I think we will have other issues with these projection problems.

[HamishB]

.. parts of georef.c will need to be replaced anyway due to license issues. hmmm... I just noticed that the header lists 3 authors, one of which is gpstrans which I know to be DFSG compliant. So which of those three authors has applied the non-commercial clause, and was that added while the opencpn code was in CVS? (so easily backed out)

What's MacGPS's license?

I'm assuming that the non-commercial part is from Carsten, maybe he could be convinced to relicense as GPL, but really I think the code should be replaced by the gold standard: PROJ.4 and/or libproj. (mainly proj.4 but Gerald's new version of libproj has a geodesic.c for great circle calcs which is the best out there). proj4 has his old geod.c.
PROJ.4
Gerald's libproj library


Hamish

[PjotrC]

Originally Posted by HamishB

could you further describe what we are looking at there and what is wrong? (I don't know the software well)
130m looks shorter in the x than in the y?

what latitude are you at, what's the map scale?

I'm at latitude 54 N. I think that the difference in measurement in exactly 270 bearing differs too much from that of 283. The bearing 001 looks OK. With longer distances, say 5 miles, it is not noticeable, but even before, looking at tracks in close zoom, I had the impression of some unexplained differences between the chart and textual position representation.

Therefore my vague request for similar experience.

True scale is 800, zoom 14x (overzoom), but it is necessary, if anchor watch and tracking should be useful. In this situation, I am almost not interested in the chart, just the behaviour of the boat at anchor.

Piotr

[HamishB]

hmmm. that non-commercial bit came from gpstrans after all, and gpstrans is based on an old+free version of the (now) commercial MacGPS. source code for old versions of both can be found at
Peter Bennett's GPS and NMEA Site - Garmin Support

grepping in the macgps45.zip source I don't find the word commercial anywhere.

shrug.
replace the whole thing with proj4 & be done with it.


Hamish
(cc source forge report)