S57 Error when Prepare All Enc

[rgleason]

The opencpn.log has errors.
What should be done to correct these?
Using 4.99.1408

[bdbcat]

Rick...


If OCPN is precessing the update correctly, then the problem is with the update file itself, coming from NOAA. If not, then we have a bug.



I will investigate this particular update.


Thanks
Dave

[rgleason]

I've turned off all the extra opengl options and the card is an Intel HD Graphics 3000, but I can set it up to use the NVIDIA Geforce 540M graphics card and it will do better I think... if opengl is part of the problem...


I wonder if I should be removing some of the files from Noaa. Before uploading the log file I had updated the USCG-01 Rnc and Enc folders and saved my opencpn.ini file to a backup and deleted the opencpn.ini to make a completely new and good version. Yet it still was crashing.


Sent a few crash reports. Wanted to get this fixed so I could work on plugins.

[bdbcat]

Rick...


Not clear from the log.


Are you crashing while trying to update these NOAA ENC charts?
Dave

[bcn]

These errors happen when some intermediate update file is broken or does not exist. Gaps in the update chain. Not crashing OpenCPN.

Eliminating the update files/directories (... .008, .007, .006) backwards one-by-one will give you the update set that crashes the sequence. The error message indicates already the probable culprit.

The Hydrographical Offices do check their sets usually well, but...
Another source of errors like this is a broken download. First thing to try is to download again and retry install.

[rgleason]

No.


I've used chartdownloader to update USCG01 for rnc and enc (as well as a number of other CG regions, then saved opencpn.ini to a backup file and removed opencpn.ini, started Opencpn 4.99.1408 again and then set up the chart directories and several chart subdirectories for charts/ENC/CG01 and charts/RNC/CG01 and then respective chart groups. Trying to keep it simple.


Then the crashes seem to continue, when I zooom in and out etc.
I am going to use chart downloader to check those downloads next.


Thanks

[rgleason]

I've used chartdownloader again on CG01 enc and rnc and have removed some older large scale east coast atlantic charts from the chart directories.



Also there were several Win10 updates both before this started happening and more recently last night. This is a Win Surface Pro3 so I was in error about changing graphics cards. It is some Intel Graphics Display, pretty high resolution.


Anyway when I update the charts, Force a Full DB Scan, then adjust all the chart settings in each canvas, it seems to be working without crashing Opencpn.



Then when I try to Prepare all enc I get "S57 Cell Update Failed Enc features may be inaccurate or incomplete. Check your opencpn.log file." each time at the same file.



This no longer seems to be causing crashes, when I zoom in and out, etc.



I attach a screenshot and another log file in case that will help, but I think I can start testing plugins now because it seems to be more stable even with the ENC Senc error.

[bcn]

Rick,

how does the directory structure of the chart where the process stalls look like? US5MA27M.000 from your screen shot.

Can you post a picture showing the directory with the base cell and the update cells in their directory? Something like in the attached screen shot..

[jstevens]

Here are screen captures of the directories on my system for the corrupted charts.

US4MD40: https://www.screencast.com/t/auwenFWyIn
US4NC11M: https://www.screencast.com/t/uyGe7nPvRs
US5MA27: https://www.screencast.com/t/GmzgnFk1M

I previously posted this issue as SENC Error. It does appear that there are corrupted updated files for each of these charts. I haven't contacted NOAA as yet, making sure it wasn't a problem with my system.

John

[bcn]

So at least there are no missing intermediate updates in the single file sets.

The next step here would to throw out the latest update, for example the .005 in the US5MA.. case and try again, until we have the offending one.
Which would have to be dissected in order to know what fails on which side.

[jstevens]

I eliminated update .005 and the problem cleared for US5MA27M.

[rgleason]

Is the latest update in this set US5MA27M.008?
When you say "toss out" what does that mean? Delete it?
Would you delete US5MA27M.008 now?


Also the error file says another process is using the file. Do we know what process that is and can I turn that process off, so that the offending file is automatically deleted as it seems that is what is required?

[jstevens]

Eliminated updates .005 and .006 for US4MD40M and problem cleared.

[bcn]

US5MA27M.000 is the base cell and 001, 002 etc. are the updates.
A new edition will be called .000 again.

008 is the latest one in your example. If you delete it and then the process does run smoothly, you got the offending data set.

And btw.: the txt files shown are observations like those from raster or paper charts. The beta 4.99 will list .txt or other additional material from the cells when querying an object. Giving access to them.

[jstevens]

I spoke too soon for US4MD40M. I still get the error, but now the chart is displayed by OpenCPN. Previously the chart was not displayed. I deleted update .004 and that chart appears to process normally.

John