Weather Routing

[stelian]

Originally Posted by boat_alexandra

So no regression correct?

Correct.

[DavidBurch]

I updated to latest 1.11 on PC, and it loads and runs, but will not compute, saying no data, but there are data. Also, maybe a main hint: the Preferences in the plugin list will not activate. I think i am overlooking something basic but can't figure it out. And suggestions will be much appreciated. It was working before i did the update. Thanks.

[Gilletarom]

Hello everyone,

I was away this week and participated at least in OpenCPN developments.

I read, with some pride, that my tests showed WR's astonishing ability to use boundarie. I regret not being able to convince you a year ago of this capacity, because it was more than a year since I understood that.

I just went through the last exchanges about WR. For a long time, now, for example between Cherbourg and Cowes, I impose on OpenCPN and WR, to use a timestamp of 10 minutes and often even 5 minutes. It will not come to me to remove the boundaries. With gshhs, weather-routes can not be prevented from passing too close to dangerous areas.

On the one hand, I thank Did-g very much for the improvements he has made to WR in terms of speed of calculation.

On the other hand, Did-g has created a formidable tool capable of creating, automatically, much more precise boundaries than mine. These boundaries, created by this tool, make it possible to guarantee a level of safety of weather-routes much higher than that of the gshhs when navigating in dangerous zones.

Currently, I consider that the gshhs, even the latest gshhs that have been installed, do not allow to create weather-routes guaranteeing the absence of risk of navigation. And therefore, we absolutely need the boundaries.
In a nutshell, we need gshhs only in areas that are not properly mapped by S57 (or S63, or oeSENC) maps.

I know that most weather-routing software claims to create "quickly", in a few seconds sometimes, weather-road, but none provides so much precision, so safety, in the quality of weather-routes
But, safety at sea is the "raison d'être" of OpenCPN, I think it is better to let WR take a few minutes to create a weather-route guaranteeing the safety of the passages borrowed, rather than want to calculate at all costs very, very quickly, weather-routes that pass too close to dangerous areas.

[kevinvh]

Originally Posted by Gilletarom

Hello everyone,

I was away this week and participated at least in OpenCPN developments.

I read, with some pride, that my tests showed WR's astonishing ability to use boundarie. I regret not being able to convince you a year ago of this capacity, because it was more than a year since I understood that.

I just went through the last exchanges about WR. For a long time, now, for example between Cherbourg and Cowes, I impose on OpenCPN and WR, to use a timestamp of 10 minutes and often even 5 minutes. It will not come to me to remove the boundaries. With gshhs, weather-routes can not be prevented from passing too close to dangerous areas.

On the one hand, I thank Did-g very much for the improvements he has made to WR in terms of speed of calculation.

On the other hand, Did-g has created a formidable tool capable of creating, automatically, much more precise boundaries than mine. These boundaries, created by this tool, make it possible to guarantee a level of safety of weather-routes much higher than that of the gshhs when navigating in dangerous zones.

Currently, I consider that the gshhs, even the latest gshhs that have been installed, do not allow to create weather-routes guaranteeing the absence of risk of navigation. And therefore, we absolutely need the boundaries.
In a nutshell, we need gshhs only in areas that are not properly mapped by S57 (or S63, or oeSENC) maps.

I know that most weather-routing software claims to create "quickly", in a few seconds sometimes, weather-road, but none provides so much precision, so safety, in the quality of weather-routes
But, safety at sea is the "raison d'être" of OpenCPN, I think it is better to let WR take a few minutes to create a weather-route guaranteeing the safety of the passages borrowed, rather than want to calculate at all costs very, very quickly, weather-routes that pass too close to dangerous areas.

Hi Gillaterom,

I have searched the Help files and can't find the answer I am looking for. With the WR PI, does it only look at the GSHHS basemaps when calculating a WR close to land, or do detailed navigation charts installed with OpenCPN for that area, take precedence over GSHHS?

[Gilletarom]

Hello Kevinh,

Originally Posted by kevinvh

Hi Gillaterom,
I have searched the Help files and can't find the answer I am looking for. With the WR PI, does it only look at the GSHHS basemaps when calculating a WR close to land, or do detailed navigation charts installed with OpenCPN for that area, take precedence over GSHHS?

WR looks at two things:
- The gshhs, provided to check the ad-hoc boxes in the definition of the search configuration
- The boundaries provided, also, check the ad-hoc box.

WR does not look at S57, S63, oeSENc charts or raster charts.

[kevinvh]

Originally Posted by Gilletarom

Hello Kevinh,


WR looks at two things:
- The gshhs, provided to check the ad-hoc boxes in the definition of the search configuration
- The boundaries provided, also, check the ad-hoc box.

WR does not look at S57, S63, oeSENc charts or raster charts.

Thanks

[bcn]

In other words: WR does not know about vector chart objects.
Shoals, wrecks, hazards, TSS.....

As of today.

[rgleason]

Language file update
https://github.com/rgleason/weather_...es/tag/v1.11.1

[did-g]

Originally Posted by boat_alexandra

So no regression correct?

I have yet to attempt optimizing boundaries.

When adding OD points I've introduced a speed regression.
Once removed routing with boundaries tests is nearly as fast as with land testing only even with files nearly ten time bigger than the one posted by Gilletarom , as a matter of fact with the right inclusion boundary it's faster.

[martinworswick]

Could this plugin receive and decode the noaa satellite images if a qfh antenna was attached?
https://www.wesacom.de/mx137_e.htm

[seandepagnier]

Originally Posted by martinworswick

Could this plugin receive and decode the noaa satellite images if a qfh antenna was attached?
https://www.wesacom.de/mx137_e.htm

This thread is for weather routing, your question is about the weatherfax plugin, and the answer is no.

[seandepagnier]

Originally Posted by did-g

When adding OD points I've introduced a speed regression.
Once removed routing with boundaries tests is nearly as fast as with land testing only even with files nearly ten time bigger than the one posted by Gilletarom , as a matter of fact with the right inclusion boundary it's faster.

Unless you have many boundaries, probably hundreds or more. There is no tree to optimize them, instead every boundary must be tried.

I'm not sure what your regression referring to OD points is.

[martinworswick]

Originally Posted by boat_alexandra

This thread is for weather routing, your question is about the weatherfax plugin, and the answer is no.

My mistake,clicked on the wrong thread title!

[did-g]

Originally Posted by boat_alexandra

Unless you have many boundaries, probably hundreds or more. There is no tree to optimize them, instead every boundary must be tried.

I'm not sure what your regression referring to OD points is.

Now that we have SQLITE requirement maybe using in memory DB with or without RTREE could make it faster.

For OD points it's the way data are stored in draw , there's no separate list or index for points in boundaries and isolated points.
When testing if a route crosses OD points you have to test every points in boundaries and isolated for the 'isolated condition' and it's time consuming, even if you have no OD points.

With Gilletarom xml data (lot of points in boundaries and very little isolated OD points), removing the OD points test make computing a routing 7 time faster, as measured on my box.

[kevinvh]

Originally Posted by Gilletarom

Hello everyone,

I was away this week and participated at least in OpenCPN developments.

I read, with some pride, that my tests showed WR's astonishing ability to use boundarie. I regret not being able to convince you a year ago of this capacity, because it was more than a year since I understood that.

I just went through the last exchanges about WR. For a long time, now, for example between Cherbourg and Cowes, I impose on OpenCPN and WR, to use a timestamp of 10 minutes and often even 5 minutes. It will not come to me to remove the boundaries. With gshhs, weather-routes can not be prevented from passing too close to dangerous areas.

On the one hand, I thank Did-g very much for the improvements he has made to WR in terms of speed of calculation.

On the other hand, Did-g has created a formidable tool capable of creating, automatically, much more precise boundaries than mine. These boundaries, created by this tool, make it possible to guarantee a level of safety of weather-routes much higher than that of the gshhs when navigating in dangerous zones.

Currently, I consider that the gshhs, even the latest gshhs that have been installed, do not allow to create weather-routes guaranteeing the absence of risk of navigation. And therefore, we absolutely need the boundaries.
In a nutshell, we need gshhs only in areas that are not properly mapped by S57 (or S63, or oeSENC) maps.

I know that most weather-routing software claims to create "quickly", in a few seconds sometimes, weather-road, but none provides so much precision, so safety, in the quality of weather-routes
But, safety at sea is the "raison d'être" of OpenCPN, I think it is better to let WR take a few minutes to create a weather-route guaranteeing the safety of the passages borrowed, rather than want to calculate at all costs very, very quickly, weather-routes that pass too close to dangerous areas.

Hi Gilletarom,

Not sure if this is the place to raise this, but I think there is a small problem with the OCPN Draw PI. I am using a MacBook Pro, OpenCPN 4.8.2 and OCPN Draw 1.4. I cannot turn off the Boundary Point Icons after unchecking "Show Boundary Point Icons" in the Boundary Properties window. I would like to turn the icons off as they clutter the screen when I zoom out.