Inland ENCs

[yachtman]

Originally Posted by nohal

yachtsman...
Not enough info. This matter is very sensitive to exactness. Please at least tell us about which charts are you talking and post the content of object query or at least exact coordinates so that we know what are you talking about. The object names are case sensitive and if it's "Lndare" on some chart we sure have no rule to display it (yet). Ienc version 2 also is a bit misleading - there are 2.0, 2.1, 2.2 and recently 2.3
BTW the main difference between iENC and ENC is that the inland objects are lowercase.

Pavel

dutch 1R5YS889.000 Ienc ver2.0 does not display other than
Feature Class: LNDRGN and
Feature Class: DAYMAR




dutch 1R5YS878.000 Ienc ver2.0 does display correctly

with Feature Class: LNDARE

Feature Class: FAIRWY
Feature Class: DEPARE
Feature Class: SLCONS
Class: LNDRGN and Feature Class: DAYMAR



for instance I understand a km marker in the riverbed doesnt show because the rule dismar is incorrect in the lookup table in some way. But LNDARE is there and displays in many tiles. I take it I need to study the syntax info more.

David

[nohal]

David...
I suppose that the first cell should be 1R5YS899, but it does not matter, the problem is the same as with all the other broken Dutch cells. They contain tons of errors - If you want to see, get CARIS Easy View | Free CARIS viewer for 2D and 3D data ...
Still EasyView and SeeMyEnc display them in a way that looks more correct. PolarView does not display them at all and OpenCPN is somewhere in between.
In this case it's not a rules problem, it's about the chart's data. And not very surprisingly the viewers using internally the same technology like the tools used to create these charts display them better...
The question is: Is it us who should try to change OpenCPN to display more of those charts or should Dutch government fix those charts?

Any incorrectly rasterized dismar location I could have a look at to check our rule?

Pavel

[yachtman]

Originally Posted by nohal

David...
I suppose that the first cell should be 1R5YS899, but it does not matter, the problem is the same as with all the other broken Dutch cells. They contain tons of errors - If you want to see, get CARIS Easy View | Free CARIS viewer for 2D and 3D data ...
Still EasyView and SeeMyEnc display them in a way that looks more correct. PolarView does not display them at all and OpenCPN is somewhere in between.
In this case it's not a rules problem, it's about the chart's data. And not very surprisingly the viewers using internally the same technology like the tools used to create these charts display them better...
The question is: Is it us who should try to change OpenCPN to display more of those charts or should Dutch government fix those charts?

Any incorrectly rasterized dismar location I could have a look at to check our rule?

Pavel

thanks for your reply, downloaded caris & I see what you mean!!


On 1RYS878 a missing dismar is at 51.57.1657 N 005 56.8102 E , Km 879 near ijsselkop, bottom middle of the chart tile, middle river off the groyne with the ski and jet ski permitted board.

[nohal]

Originally Posted by yachtman

On 1RYS878 a missing dismar is at 51.57.1657 N 005 56.8102 E , Km 879 near ijsselkop, bottom middle of the chart tile, middle river off the groyne with the ski and jet ski permitted board.

Interesting...
mine looks like the attached, which I suppose is correct. What's your settings on the Vector Charts tab in the Toolbox?

Pavel

[yachtman]

Originally Posted by nohal

Interesting...
mine looks like the attached, which I suppose is correct. What's your settings on the Vector Charts tab in the Toolbox?

Pavel

Doh !!! wasnt on mariners standard!

Just when I thought I was being useful.

Any chance of making the numbers of the Km larger as they figure heavily in Rhine navigation, particularly when going up river (in my view)

[gaea]

Just saw the German Danube Chart, from Lock Jochenstein in upstream direction.
Must say that it looks much better than the other inland charts I have seen before.
Almost usable...
Friday I will sail there and test it real time.

I agree with yachtman that visability of the km signs is realy important.


Gerard

[sailorF54]

Originally Posted by gaea

I agree with yachtman that visability of the km signs is realy important.


Gerard

As a stop gap measure, you might hack the S52 file the way it is described in post #4 here
http://www.cruisersforum.com/forums/...tml#post710829
looking for the dozen or so CATDIS entries and
changing the '14106' strings to '14112' when it occurs
(if too large for your taste, try '14110')

See snapshot of a random stretch of the Rhine river

(I haven't had time to find which ones are important for distance marks)

[Zoef]

Originally Posted by cagney

Dave

Elektronischer Wasserstraßen-Informationsservice - [Inland-ENC Dateien] also published on the wiki.
A preliminary test indicates that these charts don't quilt normally.

So.....now we are waiting for France to publish free inland charts.
Are there any plans for this at all?

Thomas

I'm sure there are plans but France being France it could take another 10 years. There are simply too many vested interests.
However, from the VNF site you can download a POI file with all the sluices in France and after that GE2Kap is your friend.
VNF a 20 ans - Voies navigables de France
On the same site you find high-quality pdf map files of the French inland waterways.

VNF a 20 ans - Voies navigables de France
Use the latest Adobe Reader (X); Foxit has problems with the files.



Klaas

[gaea]

Just tested some German charts from Danube, Main Donau channel, Main and Rhine by actually sailing there.
Overall quite usable, not bad at all.
Text like city names are shown correctly and only once(!).(unlike Hungarian chart)
Buoy's are shown like buoy's,
Exept on the Rhine sailing downstream passing km 560 (Ehrenthal) buoys are not shown as buoys anymore, but as squares and triangles...
Just saw that charts date goes from 2010 to 2006 on this point...
indeed the best inland charts I've seen so far

[yachtman]

The Dutch charts have been updated recently, fixing the display of broken tiles I have looked at so far, but not all, and broken some that worked before !



ftp://www.risserver.nl/ENC/

I downloaded totaal .zip

[Zoef]

OpenCPN crashed consistently after getting in a loop trying to load (broken?) tiles. The good thing is that the info in the tiles is very detailed.

[nohal]

Zoef...
Could you tell us which tiles are causing the crashes for you? It does not matter whether the tile's ok or not, but we should definitely not die trying to load it...

Pavel

[Zoef]

Pawel,


I'm using OpenCPN 2.4.628; the ECN file is 320Mb and contains 158 tiles. For testing I used a laptop (T2050 cpu, 2Gb ram) and a desktop (E6850 cpu, 8Gb ram).
As long as I don't enable chart quilting things are OK. However, when I enable chart quilting on the laptop, more and more tiles are opened; OpenCPN is inaccessible and when 20 or so tiles are opened OpenCPN simply closes down.
If I do the same on the desktop OpenCPN doesn't crash. A number of tiles (seemingly randomly distributed) are accessible and the others don't show at all or only 1 layer (e.g names of cities) is displayed.


Klaas

[nohal]

Klaas,
looks like you could try to play with the following settings introduced in 2.4.606:

[Settings]
MEMCacheLimit=xxx

where xxx is memory use target in Mbytes. Overrides NCacheLimit below.

--or--
b. Open chart limit. This is the default mode under linux, and the default value is 20 open charts at any one time.
Modify this limit by the following:

[Settings]
NCacheLimit=y

where y is the maximum number of simultaneously open charts.

If you do not enter any memory management specification in the config file, the following defaults apply:
a. Linux... behavior is the same as previous versions, Open chart limit is 20 charts.
b. Windows...Application memory limit target is used. Target limit is 50% of available physical RAM.

Pavel

[Zoef]

Pavel,
It is definitely a memory problem. When I opened some area's ( Ijsselmeer + Randmeren, 25 Mb) instead of the "totaal" with 312Mb everything worked fine. I can live with this.
With 8Gb of installed ram I expected that OpenCPN could use half of that (Win7, not Linux).
Instead, the mem use maxed out at about 700Mb with "totaal". With 2 area's opened, 140Mb was used, which is what I have with cm93.
Thank you for the help.
Klaas