Here is OpenCPN
Version 2.1.0 Beta 328
1. Rework Toolbar to allow improved dusk/night color control.
2. Correct chart directory fault for non-existent directories.
3. Add green conical(nun) buoy symbol to Presentation Library.
4. Improve dusk/night color schemes for controls and menus.
5. Correct flawed logic on chart dir scan after charts
have been moved on disk.
6. Improve BSB
chart georeferencing algorithms, accounting for projection type. Thanks, Marco...
7. Update chart database format to V1.6, including auto-update of previous versions.
8. Expand Right-Click context menus.
9. Correct Makefile.am to use $localedir variable for language file installations.
10. Implement Chart Quilting, F9 to toggle on/off.
11. Correct some CM93 cell offset calculations relative to WGS84.
12. Correct for upper/lower file naming conventions for charts
, esp. cm93 dictionary
13. Improve accuracy of Distance/Bearing calculations for nearly E/W bearings.
14. Implement "Course-Up" Chart/Quilt rotation mode.
15. Improve AIS
graphics, add AIS
target type 21 (AtoN), add AIS toggle tool, etc. Thanks PjotrC.
16. Add simple auto logging function. Thanks PjotrC again.
17. Add optional Ships Bells sound. Thanks PjotrC.
18. Improve geodesic(Great Circle) calculations to account for elliptic earth model.
19. Correct fault in Route/EndRoute menu with empty route
20. Correct chart startup ColorScheme as read from config file.
21. Improve RouteProp and GPX track import
usability, thanks PjotrC
22. Correct wxSockets problem leading to poor performance or hangs with gpsd and S57 SENC creation.
23. Implement Anchor
Watch, thanks PjotrC
24. Add Portugese translation, update Spanish translation.
25. Increase default MSW COMport numbers to 32, adjustable by configuration file.
26. Correct Makefile.am to honor --prefix option in configure.
27. Cleanup some source files to address licensing issues (georef.c)
28, Improve Track rendering, allow disjoint track import/export. Thanks PjotrC
29, Integrate RouteManagerDialog. Thanks, Anders.
Release Beta 2.1.0 Build 328
Lots to talk about here.
Hopefully this new thread will pull together open discussions from a number of other threads in a trackable way.
Most of the changes you will recognize, or are self-evident.
charts will use embedded georef polynomial equations if provided (BSB V3+). Otherwise, a stable low order polynomial derived from projections stated in BSB header is used, following the concepts developed by GPSMarco. Seems much better to me for Merc and TMerc charts I've tested. The code really cleaned up nicely....
7. Necessary for quilting. Update occurs when you visit Toolbox->Charts page the first time.
10. The big one. Quilting is expensive both in memory and cycles. I often see quilts composed of up to 12 charts, all of which need to be open and ready to render at the same time. Consequently, there are some subjective constants buried in the code relating to how soon (on a zoom-in, say) to begin rendering larger scale charts in the quilt. I have set the constants to roughly match the behavior of another ECS which I respect. We may need a slider control for this parameter in Release. It's all about performance. Also, note that we don't allow skewed charts (as seen in BSB/NOAA Region6 ICW
, for instance) into the quilt. Deskewing these charts into a blittable format dynamically is unwieldy and slow. Also consistent with other ECS products. Finally, TMerc chart quilting is untested. I have not enough overlapping TMerc charts, I'm afraid.
14. Course Up Mode: This code uses a fixed 15 second COG averaging/update period for the rotation angle. A compromise, again relating to performance and screen
jitter effects. This mode is only testable at sea or with a GPS
We need to add a compass
rose graphic to the display to indicate North in this mode, I think. Contributors?
22. The recently discussed hang on SENC creation. Solved
27. Relates to 18. New Geodesic calculations come from Proj4, license
OK. Datum and other stuff from public domain sources. There are possible license
exposures still. There is more work to do here if we want to use the external PROJ4 library for the various projections. Separate discussion is necessary for this whole issue as relates to debian compliance....
29. Integrated RouteManagerDialog substantially as provided, with tweaks from Psyches and me to support quilting. I've done minimal functional testing on this, but decided to include it in this Beta to consolidate the testing/review effort.
This is a lot of new code. Much refactoring for quilting/rotation. Much new real code for georef algorithms, AIS, Routes, etc... Lots of contributed patches, all of which needed line-by-line merging with my new stuff.
Also, lots of the early comments and suggestions for improvements to 1.3.6 fell off the list. If your favorite idea/bug didn't make the list this time, I apologize. I'd be delighted to hear from you again, to keep it on the big ToDo list
I expect at least some regression problems with this code. That is, things that used to work are now broken, perhaps in ways that provoke crashes. Also, some new functions have not really been stretched very hard in my alpha testing.
But it does build and run, so we release early and often. There is no real substitute for the variety of environments seen in our test group.
A caveat is necessary.
If you use 1.3.6 for real at-sea navigating, I'd suggest you keep a clean copy of it available to use daily as we work through the first few Betas of 2.1.0. I do....
CVS and git....
This Beta will use the current
CVS schema for two reasons:
1. Minimize the development deltas a bit, in order to reduce risk of total breakage.
2. I personally need time to learn git.
I expect to transition to git in the near future as time permits. Seems like a good thing....
Meanwhile, CVS is updated for Build 328.
So, here we go again. I'm excited by this new version, and hope you are too. Beta testing is fun!
I think we can reasonably plan for a production release of OpenCPN
Version 2.1.0 in a month or so.
Thanks in advance for all you do...