OpenCPN Version 2.4 Beta Build 423

[cagney]

Dave/Pavel

It seems like a lot of top marks are just gone in CM93 and replaced with A square with a "?".
Looking at Dry Tortugas with a US ENC

Using CM93

Details of the NW mark


Thomas

[PjotrC]

A small fix, that didn't make it for previous release. It removes the (meaningless) tide information for past waypoints on an active route.

[HappySeagull]

I have the buoys,no lighthouse!
If we could just all put our cm93s in a kennel together so they could interbreed...

Nohal,F11, works perfect on my XPsp1(sorry,apologies I missed it in the OP!)

[nohal]

Thomas,
might be - the object query shows that the shape is just in INFORM parameter as a text, so we read "some red topmark" which we don't symbolize (actually we do as "some topmark", that's why the symbol used) - the object query on ENC will have the shape parameter set (don't have the chart here right now to check, but am pretty sure)
IMHO the best we can do is make a red general topmark symbol (red square with a question mark) as trying to take care of the INFORMs is a way to hell - they are too entropic across the globe.

Pavel

[nohal]

For all the CM93 editions floating around I can of course say just: NO COMMENT

[cagney]

Originally Posted by nohal

Thomas,
might be - the object query shows that the shape is just in INFORM parameter as a text, so we read "some red topmark" which we don't symbolize (actually we do as "some topmark", that's why the symbol used) - the object query on ENC will have the shape parameter set (don't have the chart here right now to check, but am pretty sure)
IMHO the best we can do is make a red general topmark symbol (red square with a question mark) as trying to take care of the INFORMs is a way to hell - they are too entropic across the globe.

Pavel

Pavel you're right about the shape parameter in ENC.
I think that a red general top mark in system B should be a triangle with top up, in system A it should be a square. The meta info contains MARSYS.

Happy!
Your breeding suggestion may lead to a worst case scenario...

Thomas

[sinbad7]

Running OpenCPN v 412 under Windows Vista is too unstabel,crashes when accessing the most usual features. I have reverted back to v. 324 until the problems have been resolved.

Tore

[bdbcat]

Tore...

Have you tried 423? No big problems reported yet.

Dave

[sinbad7]

Dave..

I'm sorry, I meant v 423.. Have frequent crashes at the most unexpected places.
Both using the layers function as well as Show chart outlines etc.

Tore

[bdbcat]

Tore....

We have no crash reports concerning these areas on 423. The only crash report we have concerns very small scale quilted underzooming in locations without any charts present at the viewport. Takes some effort to reproduce...

Your specific crash cases and screenshots would be most helpful.

Thanks
Dave

[PjotrC]

For Alands Archipelago (60 30N, 22 00E) the 1.3.6 is still the version of choice, I think... No crashes, but performance in zoom/pan... Also, the readability of chart is better - could the thick safety depth contour and buoy topmarks be opted out perhaps or matched to the zoom level?

It is good that 1.3.6 can coexist fully with 2.4. - one can have the best of both worlds!

On the fresh 423 installation I had one crash that I can not reproduce, in a situation that reminded me of previous versions - when (simulated) underway, slowly moving the chart in very small increments (follow the boat) for a period of time. It was going W of Curacao this time.

Piotr

[nohal]

Originally Posted by PjotrC

For Alands Archipelago (60 30N, 22 00E) the 1.3.6 is still the version of choice, I think... No crashes, but performance in zoom/pan... Also, the readability of chart is better - could the thick safety depth contour and buoy topmarks be opted out perhaps or matched to the zoom level?

Zoomed not, opted out very easily - Toolbox -> Vector Charts -> uncheck TOPMAR and DEPCNT in the listbox on the left. The only drawback is that you loose all the DEPCNTs, not just the "fat" one.

Pavel

[bdbcat]

Hello testers....

423 status:
The Beta 423 core is very close to releasable. There will be several subsystem bug fixes as described in Flyspray, and some new UI candy coming soon. But the base engine is, IMHO, stable and complete.

However, there are reports of intermittent and difficult to reproduce crashes.
That's not good.

I'd like to tighten up the testing here, specifically looking for crash cases in normal operation. The exposed areas seem to be related to quilted pans and zooms, with perhaps vector charts being more susceptible. Just an impression...

I plan to release OCPN 2.5 in June, so we have a month or so to iron this out.

I really appreciate all your efforts in this regression testing. I know it can be tedious at times....

Thanks
Dave

[sailorF54]

Originally Posted by bdbcat

specifically looking for crash cases in normal operation. The exposed areas seem to be related to quilted pans and zooms, with perhaps vector charts being more susceptible. Just an impression...
Thanks
Dave

I've had a few crashes while extensively mouse dragging CM93 charts with quilting on. It has never happened with quilting off. (Latest Vista)
Sorry I can't be more specific for the time being...

Added:
I've managed to get a reproductible crash condition on start-up with 'quilting' on as soon as I drag the chart

I get an 'unhandled exception'

This condition disappears if I uncheck the quilting option

End of log
...
14:55:31: Initializing Chart C:\Users\François\Documents\CM93_2010_5\00300000\A \00300000.A
14:55:31: CM93Composite Chart Root is C:\Users\François\Documents\CM93_2010_5\
14:55:31: Loaded CM93 Dictionary from C:\Users\François\Documents\CM93_2010_5\
14:55:31: Loading CM93 cell C:\Users\François\Documents\CM93_2010_5\03901020/E/04161068.E
14:55:31: Loading CM93 cell C:\Users\François\Documents\CM93_2010_5\03901020/E/04161069.E
14:55:31: Loading CM93 cell C:\Users\François\Documents\CM93_2010_5\03901020/E/04161070.E
14:55:31: Getting .gpx layer files from: C:\Program Files\OpenCPN\layers
14:55:32: Loading CM93 cell C:\Users\François\Documents\CM93_2010_5\03901020/D/04141068.D
14:55:32: SetCurrentViewPort called by OpenCPN
14:55:32: Following the viewport to Lat=48.805100, Lon=-3.441600, H=0.245287, W=0.710439, R=0.000000 requested
14:55:32: Not detaching, asked to catch up
14:55:32: Rotation translated to 0.000000 degrees
14:55:32: LOGBOOK: 2011-04-28 12:55:32 UTC OFF: Lat 48.80510 Lon -3.44160
14:55:32: opencpn::MyApp exiting cleanly...

----
Don'k know if it is relevant

On previous occasions, I had the following error in the log

13:39:59: Getting .gpx layer files from: C:\Program Files\OpenCPN\layers
13:39:59: Error: Can not enumerate files in directory 'C:\Program Files\OpenCPN\layers' (error 3: le chemin d'accès spécifié est introuvable.)

I then created an empty folder 'layers' in C:\Program Files\OpenCPN\layers
-----

[bdbcat]

SailorF54....

Regarding the reproducible crash:

This is very relevant.

Could you please attach a copy of your config file that causes the crash on first drag?

Thanks
Dave