Avalon .10

[R3d2]

TWD and TWA


Downloaded latest weather forecasts from ~ 196h and this appeared.
ZEZO everything is fine.
Restarted browser.
Restarted program.
Restarted windows 10 - waiting for NMEA latest position

[jef]

Having same trouble here, seems just a bad re-computation of TWD, because DASHBOARD only send true wind angle (TWA), so obviously i mess something doing TWD = TWA - SOG !!!!


Think this is a DASHBOARD issue, from my dev console this is what i read from DASHBOARD :

COMPUTE-ROUTE:: RMC: lat=-49.80863833333333 lon=86.70728666666668 sog=20.2 cog=81.2
COMPUTE-ROUTE:: MWV: TWS=20.31 TWA=307.96

This is not important for router so trust your isochrones and route, will investigate to understood where is that mess !

[R3d2]

Restarted Zezo and again the same problem appeared.

[jef]

If you use VRDASHBOARD 3.14.2, then you must use 1.10 original.
VRDASHBOARD 3.14.1 send TWD instead of TWA, this was a mistake from VRDASHBOARD.
And corrected nack in version 3.14.2, so the original AvalonVirtual 1.10 is right now.

Bye

[R3d2]

If you use VRDASHBOARD 3.14.2, then you must use 1.10 original.
VRDASHBOARD 3.14.1 send TWD instead of TWA, this was a mistake from VRDASHBOARD.
And corrected nack in version 3.14.2, so the original AvalonVirtual 1.10 is right now.

Bye

Fun is that, that I'm using 3.14.2 and 1.10 AvalonVirtual.jar 8.12.2020 07:47, 31 471 459 bytes

At beginning was everything fine.

[R3d2]

There is some mistake. Wrong sail showed up
As in table at 18 kTs TWS 132 TWD
Code-0 foils 18.61 kTs
Heavy-gnk foils 18.66 kTs

V 1.10-1
Version 3.14.2
java version "1.8.0_271"
Windows 10 64bit

[jef]

Hi, R3D2

Just update your polars with TOXCTT, this have been updated (and no one knew that !)
Having same trouble here, just reload my VR page, and then able to have TWA 132 !!!!!! Weird
Bye

[R3d2]

It's unbelievable mistake to calculate SOG from pollars and not to put IF in calculation!!! When SOG at 33 -> 34 ->35 --> 37 kTs

Hope you fix soon!!! Moves routing to ZEZO

[R3d2]

It's unbelievable mistake to calculate SOG from pollars and not to put IF in calculation!!! When SOG at 33 -> 34 ->35 --> 37 kTs

Hope you fix soon!!! Moves routing to ZEZO

Ok. As I understood from Jef in his post on French Technical discussion, there is problem with - Filter laylines. Turned off. Seems fine.

[jef]

Hi R3D2,
Fix is coming soon. Filter LayLines is incorrect at limits, and actualy we are in limits

[R3d2]

V 10-4
Using and lost in days almoust 30min to Zezo. Calculating insane 2 pass to Cape Horn. On i9 ~ 6-7 min. Latest weather forecasts. Just only one time something changed cardinally. That's 20-30 min. lost. Overall 1 hour lost. 2 hours to first place. Something not really right on such a much movements, manouvres and we Avaloners should be in top 50. I don't see that.

[R3d2]

Real router fight. On the night the Avalon with High calculation was in front of Zezo. On morning weather forecast Zezo is in front of Avalon.
35 min difference on 2000 nm.
Tried all calculation methods - result ar Cape Horne 4:20 constantly.
Feels like calculation methods sometimes not working. Sometimes are pretty fast. Sometimes takes on i9 gen9 5 minutes

Twice restarted Win 10. On the seccond time calculation was fast. Tried High calculation. On first 24 hours Avalon was in front. Clear lines.
Small course changes and then iterations and lot of course changes. Zezo catches and overtakes on Cape Horn Avalon by 20 minutes.

[R3d2]

[Roots]

When avalon takes too long to calculate, no need to reboot windows, you can kill the java process.

[R3d2]

Killing Java process not always works well.
Last time screen hanged up. Killing process turns off the screen but problem appears again!

[R3d2]

Ok.
Before restart to destination point.

• 12/20/2020 18:58:56 Automatic 5 miles iteration 17:40 22 dec

• 12/20/2020 19:04:05 Automatic after restart 5 13:45 22dec

[R3d2]

• 12/20/2020 18:58:56 Automatic 5mils 17:40 22

• 12/20/2020 19:01:53 High 5 13:50 22

• 12/20/2020 19:04:05 Automatic restart 5 13:45 22

• 12/20/2020 19:04:11 Ultra 5 13:40 22

• 12/20/2020 19:05:05 Insane 5 13:35 22

• 12/20/2020 19:06:49 Insane 2 pass 5 13:35 22

• ZEZO 14:00 22

• 12/20/2020 19:25:03 Insane 2 pass 2miles 13:46 22

That is on 2 days. When takes longer period. Zezo wins. Avalon gets stucked and sometimes gives longer route. As route gets calculated at least 4 times a day, then we don't see much difference as a part of the luck.

As I observed something weird happening sometimes. Too long calculations, funny directions - In wrong direction, passing over dirrection. Then definitely should restart computer. It's not connected with java itself as more with library and Windows 10 speed optimization.
How is on MAC and Linux. I don't know.

[R3d2]

Made some calculations.
Interesting that Ultra promised the best result and after restart of program the time was better.

Procedure was to get as much close to time of ZEZO in coordinates and calculate correlation of route ending point as we know that to go 1 nautical minute with speed ~20 kTs will take 3 minutes.

[jef]

@R3D2,

When computing long distance routes, use INSANE 2PAss => iso are 1/83° precision (you can edit theses in Advanced Parameters)

[R3d2]

@R3D2,

When computing long distance routes, use INSANE 2PAss => iso are 1/83° precision (you can edit theses in Advanced Parameters)

I know. I already made over custom, as you see - 1/102

With this table I wished to show, that not always Avalon is good on calculations on choosing Insane 2 or more.

I made scientific experiment.

As at the start of the pacific there was clear that I should go with group at the south. Zezo offered the same. Avalon calculation Insane 2 moved to North. I tried. I and Zezo was right. I lost at least 7000 places That's why I'm frustrated! Ok. Sacrifies for science and if you wish to win prepare to loose.

Started too use Zezo then compared calculation times.
Compared custom ones as change turning points more precise.
Appeared that calculations, as they going more precise then routes getting more "longer" in some circumstances. I don't know why. I don't see calculation formula. My be is some problem with so high wind or else?