Post by iconPost by llamedos | 2009-07-21 | 20:46:08

Are you planning to support this race? I hope so as your programme is great.

commenticon 106 Comments
Post by iconPost by zezo | 2009-07-21 | 22:06:27
Yes. It requires relatively small effort (unless there is some major change like the wind update times and the interpolated polars) and I don't have other activities planned around the start.
Post by iconPost by BennyLava | 2009-07-29 | 00:10:07
Thanks zezo. Your software helped me a lot in the Volvo race. I usually ended in the top 200 (needed my nite rest, otherwise might have ended higher ;)
Post by iconPost by BennyLava | 2009-08-01 | 13:58:28
zezo; there's still a slight discrepency in the pop-up info for the boat speed. The game shows 7.1 knt, the pop-up 6.9. Otherwise a fantastic job!
Post by iconPost by zezo | 2009-08-01 | 14:16:44
Thanks. It's the case with difference in wind direction to start with. Still trying to figure out how exactly the game misinterpolates winds :-)
Post by iconPost by TSIONA | 2009-07-22 | 13:57:51
For all those using this software :
visit the donation topic in this forum, to keep the site working !
Post by iconPost by llamedos | 2009-07-26 | 23:12:11
forgot to say the new additions of pro sails and avoid tacks are great additions, looking forward to La Solitaire du Figaro 2009, thanks again and as TSIONA says visit the donation topic if you use this software.
Post by iconPost by OhLocoo | 2009-07-29 | 04:44:17
If help you

One chart of spain+france+UnitedKingdon in .NAV

http://www.4shared.com/file/121381630/634c9e1/Figaro1.html
Post by iconPost by zezo | 2009-07-29 | 20:21:39
Thanks. I'll use the coordinates together with the BPE polars before the start.
Post by iconPost by TSIONA | 2009-07-29 | 10:37:20
Thank you OhLocoo,
I will use those charts.
"See" you tomorrow.
Post by iconPost by zezo | 2009-07-30 | 15:16:40
A lot of new features in the game so it will take a day or two to get in sync.

New features:

- Finer grained wind grid - 1/4 degree.
- Morning forecast is new - possibly the correct one, but that remains to be seen
- Polar diagrams are same as BPE, but interpolated like in tdf.
Post by iconPost by MMB | 2009-07-30 | 15:19:11
Winds are not the good one, right ?

;o)
Post by iconPost by BennyLava | 2009-07-30 | 15:52:35
The obligatory Belle-Ille buoy is missing?
Post by iconPost by zezo | 2009-07-30 | 16:15:54
That's smallest problem. It's a matter of 5 minutes of work, but right now the site is mostly unusable anyway, so it does not make much difference.
Post by iconPost by BennyLava | 2009-07-30 | 16:32:21
You're right zezo, don't let you distract from the most important task...;)

Cheers!
Post by iconPost by zezo | 2009-07-30 | 16:48:40
Morning winds seem to be GFS0Z+0 instead of +6. It was +9 (incorrect) during the TDF.

This is confirmed by the file headers:

<PREVISION PREVISION="0" DATE="2009-07-30 00:00:00">

Let's see what comes out in 4 hours.
Post by iconPost by TSIONA | 2009-07-30 | 17:41:41
The Cvetan's step by step progressions are making this site more interesting than the other...
How in 4 hours ? The actual wind starts at 08:00. The 12 hours forecast of the game that is to be added and here is 20:00.
Post by iconPost by zezo | 2009-07-30 | 20:05:10
Here go the polars.

Wicked gap between jib and spi.
Post by iconPost by MMB | 2009-07-30 | 20:30:38
Could you give the polars as .csv ?

;o)
Post by iconPost by zezo | 2009-07-30 | 21:01:21
The files may not be usable directly with VRtool but should be good start:

http://zezo.org/figaro/classic.csv
http://zezo.org/figaro/pro.csv

and with best sail information included:

http://zezo.org/figaro/classic_s.csv
http://zezo.org/figaro/pro_s.csv
Post by iconPost by MMB | 2009-07-30 | 22:13:08
VRTools says "error in file " !

;o)
Post by iconPost by zezo | 2009-07-30 | 21:09:42
The forecast timestamp is seriously wrong:

<PREVISION PREVISION="0" DATE="2009-07-31 06:00:00">

also shown in the interface as 8h GMT+2
Post by iconPost by BennyLava | 2009-07-30 | 21:13:52
maybe they try to hide it from you?
Post by iconPost by zezo | 2009-07-30 | 21:22:40
I can always download the files from the game server, so no way to hide. Some people just never learn.

Current winds are indeed the tomorrow morning forecast.

I wonder if the winds in the morning will be from the period before the current forecast. Maybe we should start betting on what comes out at every wind change ;)

Anyway, good luck to all players relying on reasonable forecasts as it makes no sense building strategy with random data.
Post by iconPost by BennyLava | 2009-07-30 | 21:32:47
"Anyway, good luck to all players relying on reasonable forecasts as it makes no sense building strategy with random data."

So, does this mean that their own forecasts won't be the next winds and that we won't get the correct winds on your chart? I haven't sailed the tour de france, how did you get the wind forecasts?
Post by iconPost by zezo | 2009-07-30 | 21:44:13
The current 12h forecast is in fact 24h forecast - look at the "Wind charts from ..." in the game.

And if what comes as 0h in the morning is the same like today - 0h GMT, the difference will be significant - 18 hours and most people won't be happy.

I can display and use the proper forecast here if there is repeatable pattern, but people using the game interface itself will be totally confused.

It was the same in TDF, but the difference was smaller - +/- 3 hours, and people still complained a lot in the game forum.

The forecast is taken from the same source - NOAA.gov, but the game developers manage to screw it most of the time.

So don't rely on zezo.org displaying the same forecast as the game soon. I will implement the current game logic, but then they will probably change it again trying to fix it.
Post by iconPost by llamedos | 2009-07-30 | 22:00:31
they have just changed the winds again, what are they playing at?
Post by iconPost by zezo | 2009-07-30 | 22:10:47
C00l :-)

They got in sync with me (and the real world)

The code to convert gfs timestamps to game is somehow tricky indeed, but not something you can't get right in 1 month.

I still want to bet what comes out tomorrow:

1) GFS0Z+0h like today
2) GFS0Z+9h like in TDF
3) GFS0Z+6h (the correct one)
4) Something completely different.

I'd bet on 1 myself. Winner gets a beer if we ever meet.
Post by iconPost by BennyLava | 2009-07-30 | 22:08:34
yeah very strange???
Post by iconPost by llamedos | 2009-07-30 | 22:32:13
there is still a slight difference between their wind speeds and directions to yours but near enough
Post by iconPost by zezo | 2009-07-30 | 22:37:09
The whole degrees should be the same. I will implement the 0.25 degree grid later but it's not clear if it's based on 1 or 0.5 degree gfs data.

P.S. And I haven't applied the proper polars yet.
Post by iconPost by llamedos | 2009-07-30 | 22:55:11
yes, sorry my mistake I forgot about the new 0.25 degree grid they have implemented (I must look closer at things next time and keep my mouth shut)
Thanks for all your hard work for the rest of us.
Post by iconPost by zezo | 2009-07-30 | 23:42:34
There is nothing to be sorry about. The new grid was quite a surprise, and I rely on users for quality assurance, so it's better to spot discrepancy that is not there than to miss one that is, no matter if the bug is related to the game or the site itself.
Post by iconPost by TSIONA | 2009-07-31 | 08:37:07
Cvetan, just to make it clear (to me).
1. Is there, right now, any information displayed on the chart that can be used ?
2. After next winds change, will you be able to give us an estimation time to see the Optimizer entirely working ?

Of course, those questions have no mandatory status !
All this work of Rev.Eng. on the MP site, you are doing, is fascinating. Thank you, one more time to make us participate somehow with you to this effort, and to pursue the dreams of players to get some victory...

Ps: if somebody ready to "take" my boat this evening for 24 hours, please let me know it on this site. Thanks.
Post by iconPost by zezo | 2009-07-31 | 09:28:43
1. Not really.
2. Later today. I'll fix the polars first and that should be useful in steady winds and for general strategy by itself. Then the new wind grid will require some debugging.

Amazingly the forecast is the correct one (0Z+6)
Post by iconPost by zezo | 2009-07-31 | 10:41:56
Sails and polars should be in sync now. Please let me know if you see difference between boat speed or best sail for the same wind speed/TWA here and in the game.
Post by iconPost by BennyLava | 2009-07-31 | 11:04:17
Excuse my ignorance, but the winds on the chart are still not in sync with the game?
Post by iconPost by zezo | 2009-07-31 | 11:20:51
See the answer to llamedos above. The wind arrows with integer coordinates should already match the game. Popup data should also match if your boat happens to be within 7.5 minutes lon/lat of such integer point.
Post by iconPost by kikouic | 2009-07-31 | 11:31:36
hi! i send you some speed now i am on wind value 6.5 if you wand more tellme
Sail standard (not pro)
Wind 5.9
Spi
Twa speed
80 :1.6
90 :4.2
100 :3.8
110 :4
120 :3.9
130 :3.8
140 :3.6
Foc
40 :3.3
50 :4.1
60 :4.3
70 :4.4
80 :4.3
90 :3.6
100 :3.5
110 :3.3
120 :3.2
130 :2.9
140 :2.6

Wind 6.5
spi
Twa speed
80 :1.8
90 :4.6
100 :4.3
110 :4.5
120 :4.4
130 :4.2
140 :4
Foc
40 :3.6
50 :4.5
60 :4.6
70 :4.7
80 :4.7
90 :4.6
100 :3.8
110 :3.7
120 :3.5
130 :3.2
140 :2.8
Post by iconPost by zezo | 2009-07-31 | 11:54:48
Thanks. That's pretty close, only the jib is slightly off at 5.9, The rest is UI rounding error.

90 4.215
100 3.825
110 4.015
120 3.9325
130 3.8115
140 3.6135

40 3.329
50 4.115
60 4.23
70 4.33
80 4.225

90 4.65
100 4.3125
110 4.525
120 4.375
130 4.2
140 3.975

40 3.65
50 4.5
60 4.6
70 4.7
80 4.675
90 4.65
Post by iconPost by zezo | 2009-07-31 | 11:57:49
Hmm. The linear equation for jib/TWA60@5.9 is:
2.90+(4.30-2.90)*(5.9-4)/(6-4) = 4.23
Post by iconPost by BennyLava | 2009-07-31 | 11:46:59
Zezo; the pop-up is not quite the same:

Pop-up Game
Distance 266,3 267
Wind 5,9/128 7/126
Heading 218 231
Sail Genova Spinaker
Boat speed 4,5 5,2

The heading and boat speed are quite different!
Post by iconPost by kikouic | 2009-07-31 | 11:35:45
you are wright for wind = 5.9
Post by iconPost by TSIONA | 2009-07-31 | 11:45:03
So far so good for me concerning data appearing in green popup.
In white popup, TWA and and BS are still wrong.
You may use my username to verify.
Post by iconPost by TSIONA | 2009-07-31 | 12:00:13
Forgot to say.
For a symbolic salary i could be your QA... (laugh)
Post by iconPost by zezo | 2009-07-31 | 12:07:08
OK, just remind me to add some code to avoid double posts on refresh some day ;)
Post by iconPost by zezo | 2009-07-31 | 12:09:39
Maybe this will do the trick:
alter table forum_posts add constraint nodoub unique(user_id, thread_id, reply_to, text);
Post by iconPost by TSIONA | 2009-07-31 | 12:29:17
Already implemented ?

Double text occurs when several windows of zezo.org were simultaneously opened.
Post by iconPost by TSIONA | 2009-07-31 | 12:37:29
Just a though concerning the race.
The Thomas Rouxel's boat seems to be positionned exactly toward the right heading, like indicating the right route.
If it's true. It is interesting to know whether MP uses an Optimizer to do so...
Post by iconPost by TSIONA | 2009-07-31 | 12:38:33
The mail appears once.
Post by iconPost by zezo | 2009-07-31 | 14:18:16
I'll start migrating the site to 1/2 and later 1/4 degree wind grid soon.
Expect major disturbance in the Force in the following few hours ;-)
Post by iconPost by Belladonna | 2009-07-31 | 15:10:26
If it was about aviation, I'd said turbulances !!! :-))
Post by iconPost by kikouic | 2009-07-31 | 15:37:46
in 46N22 / 4W12 the wind is 8.1/128 not 8.1/118, 10° mistake ?
Post by iconPost by zezo | 2009-07-31 | 15:56:14
I'm still at the 1/2 degree stage, and half of the zones are a bit off. The good news is that 1/4 degrees is 4 more lines of code, but those lines require some thinking.
Post by iconPost by BennyLava | 2009-07-31 | 16:18:39
THANKS zezo for all the effort.

Just wondering do you plan to work on Sail Online (http://www.sailonline.org/) also??
Post by iconPost by zezo | 2009-07-31 | 16:35:54
If it becomes more popular. And it may become now when their ad is displayed at VR all the time ;)
Post by iconPost by BennyLava | 2009-07-31 | 17:18:36
Zezo; in my community SOL is becoming more and more popular because it's a much better designed game. Please have a look, Thanks!
Post by iconPost by MMB | 2009-07-31 | 16:33:14
Zezo President !

;o)
Post by iconPost by zezo | 2009-07-31 | 16:44:07
Stage 2 is finished. The winds should be (mostly) the same as in the game now. Let me know (with coordinates) if you notice differences. They are easily spotted for your current position, but take note of the displayed time stamp, as the game server position updates are slower than ever.

What is left now is chasing possible bugs and fixing the GUI.

The background chart needs new wind cell borders and new wind barbs at half degree (there won't be space for 1/4 degree display)
Post by iconPost by kikouic | 2009-07-31 | 16:54:01
Congratulations big boss!
Post by iconPost by Belladonna | 2009-07-31 | 17:01:32
This is really becoming a masterpiece. Chapeau Czetan !
Post by iconPost by kikouic | 2009-07-31 | 16:59:20
I does not understand what mean "1/4 degree wind grid for beta testing"
but it is true that French people are mediocre user of foreign languages
:-)
Post by iconPost by zezo | 2009-07-31 | 17:09:40
I don't like it either, but there is too little space on the fullscreen page for proper description :-)
Post by iconPost by llamedos | 2009-07-31 | 18:16:21
Things seem to be working fine, great job, many thanks.
Post by iconPost by llamedos | 2009-07-31 | 22:01:09
not sure if I've found a slight bug/mistake, the game site shows me at
5W16/46N05 with details of wind speed 11.3kt @ 216 deg but on your site it shows me at same co-ordinates but 11.3kt @ 214 deg

looks like their mistake after looking again, wind speeds either side of the above are 217kt and 211kt reading east to west and 211 kt and 222kt north to south.
Hope this all make sence.
Post by iconPost by zezo | 2009-07-31 | 22:40:42
Thanks for the observation. This one is pretty minor, but it could be worse, so it's good idea to check the data.

The grib data comes in 0.5 degree resolution, so the .25 degree squares are subject of interpolation. And that one is subject of rounding errors.

The game uses pretty weird units too. Forecast data comes in integer Km/h, but the polars are in knots, and that generates a lot of rounding errors, like the famous 20kt bug in the VOR.

There is also a +/- 0.1 kt rounding error in the current polars.

While it's possible to reverse-engineer all the quirks it takes some time and observations like this one, but it probably does not affect end result too much so I may not bother 'fixing' the difference. It could be worth for longer race but this one will be over before we find all the underwater rocks :-)
Post by iconPost by zezo | 2009-07-31 | 23:35:23
Just for the record. I'm in the same square now. Between 11.9@217 and 11.3@211 (46N 5W). The game interpolates this to 11.3@216 and my algorithm gives 11.3@214 indeed. 11.3 is rounding artefact - that's 21 km/h. 11.9 is 22 and no intermediate value is possible in the game universe. I may take a look at it tomorrow.

Original GRIB data in U/V components:

46.000 354.500 3.680000
46.000 354.500 4.800000

46.000 355.000 2.960000
46.000 355.000 4.870000

calculated speed/direction + U/V:

-5.0 46.0 211 21 2.96 4.87
-5.5 46.0 217 22 3.68 4.80

Wind direction is atan2(u,v)+180

My algorithm assumes u=(u1+u2)/2 for the interpolation.
Post by iconPost by Fuga9 | 2009-07-31 | 23:57:44
Good Job....Congratulations...!!!
Post by iconPost by zezo | 2009-08-01 | 19:42:50
I'm giving up (temporarily) the interpolation problem.

Some research shows that linear interpolation of neighbor U/V values is not the best possible solution (http://www-md.fsl.noaa.gov/eft/publications/papers/IIPS/Stu95/IIPS95StuPaper.html), but should be good enough most of the time.

There are few possible approaches but I could not think of one that will result in wind direction 216 in the example above. So the game implementation is either complex (higher order interpolation) or just buggy.

The observed difference is 1 or 2 degrees, and small speed difference sometimes. That does not affect the result too much, just take care to use the suggested TWA instead of the heading and not fall in the gap between Jib and Spi polars at TWA 85-90

Now going to pack the luggage for some real sailing next week.
Post by iconPost by llamedos | 2009-08-01 | 20:29:32
I did come across some that were out by 4 degrees today but did not log them or bother you, you've done enough already and it's great. Have a great week, you deserve it.
Post by iconPost by zezo | 2009-08-01 | 21:14:23
Thanks. I've studied 3-4 cases today with no success.

BTW the easiest workaround is importing the current weather directly from the game server. That would take less time than trying to reverse the calculations, but I still want to make the site as much independent from the game servers as possible.
Post by iconPost by Fuga9 | 2009-08-02 | 02:13:49
If u could be independent it would be a great step. I wonder...
Anyway, thanks Zezo for a great work...!!!
I do not understad how MP could blocked VrToll without blocking you..
From Brazil my regards..
Post by iconPost by zezo | 2009-08-02 | 03:17:53
It is independent with the only exception of loading the current position from the server.

Check your mail regarding the other question ;-)
Post by iconPost by OldChapZ | 2009-08-02 | 19:34:31
Omar, VRTools 's developper, could have it working the same way Cvetan is doing.
I got it working but I need the personnal "key" for "getting user" data. I assume Cvetan can only use the public "profil". What about the incertainty in computing Heading = Windir +/- Angle ? AI guess ;-) or URL I don't know about.

Great job, Cvetan. I wish you a good sailing week.
Post by iconPost by TSIONA | 2009-08-02 | 12:14:16
Did you know it was possible to get wind forecasts (matching new game requirement, 0.25 deg) at the following address :
http://www.lasolitaire.virtualregatta.com/resources/winds/meteo_-5_47.5.xml (Source: wwww.grib.us)

NB:
meteo_-5_47.5.xml can be replaced by :
-10_45, -10_50, -5_-50, -5_-45, -7.5_45, -7.5_50, -7.5_50, -7.5_47.5, -5_47.5 etc according to area coordinates.

-10 = Long 10 W
45 = Lat 45 N
Post by iconPost by BennyLava | 2009-08-02 | 15:15:41
zezo; are you around for the 2nd leg, starting August 4th at 2:00PM (GMT+2)?
Post by iconPost by zezo | 2009-08-02 | 15:21:15
I'll be away from 3th to 10t August, but tonight I'll add the destinations of leg2 and leg3 to the menu.
Post by iconPost by TSIONA | 2009-08-02 | 17:33:22
Hi Cvetan,
I am sure the boat pitch in your forthcoming real sailing will evocate thoughts concerning game and zezo.org.
So maybe it's a good idea to make you consider the following proposition now:
Since the chart presented in your site is in low resolution (because, as you said in the past, a low CPU resources VS number of players). So maybe it's a good idea to think about a way to improve the quality chart display.
As i said earlier, the right forecasts are received directly from their site including any kind of manipulation processed by MP on those files. So why not to use forecasts directly on a stand alone chart ( simple chart like in Vrtool)and to transmit there the relevant information like
TWA, position....The routing algorithm will be used from your site or from a stand alone software.
Don't think about too much !!! At the first place have a good time.
To be continued.
Best regards - Thierry
Post by iconPost by OldChapZ | 2009-08-02 | 19:56:14
Not so easy, I think. Furthermore, I'm not sure we need a better accuracy for the optimizer.
But what about a menu option to download a text file of the optimized route ?

For the server overload (wich will come as the site is wider known and used) I checked if my Internet provider allow Perl to offer a mirror. He doesn't though he uses it for own needs. Could the scripts be easily rewritten with PHP ?

[EDIT] Overload is comming soon ! See this topic on VR's forum.

Cvetan, if you want to mirror : Free Perl hosting.
(Google is [my|your|our|their] friend ;-)
Post by iconPost by zezo | 2009-08-02 | 22:57:35
Thanks.

Too bad I won't be here to watch the overlord coming (or failing to appear :-)

The main router is compiled binary written in C for efficiency and that complicates hosting solutions - it boils down to dedicated server. You can find some info on the About page and the Optimizing the optimizer post.

You have to enclose the links in [ url]...[ /url] tags manually to make them clickable.
Post by iconPost by OldChapZ | 2009-08-07 | 17:45:41
You were right : no overload coming (neither overlord ;-) )
Maybe they found the site useless ! Or dont manage to use it.


I read enough about optimization to understand I've to go back to the nursery school (I left 60years ago =:-o )

OK for the links Thks
Post by iconPost by zezo | 2009-08-02 | 22:47:03
I've entered the leg 2/3 waypoints as given by the current course file. If something changes you can always set your destination and current position from the context menu or by control/alt/shift-clicking.

See you again on Aug 10th.
Post by iconPost by BennyLava | 2009-08-02 | 23:26:22
Cheers mate! Enjoy your well deserved holidays!
Post by iconPost by Belladonna | 2009-08-04 | 14:39:45
Good winds Cvetan
Post by iconPost by Fuga9 | 2009-08-04 | 15:40:13
The software is doing very well...
Post by iconPost by Fuga9 | 2009-08-04 | 16:11:02
Thanks Cvetan....happy hollidays...:)
Post by iconPost by BandB | 2009-08-05 | 10:30:02
Hi - wonder if anyone knows why I cant get my boat to show - worked in VORG (when I found this site towards the end!). Boat name is B & B. I can get other boats to come up, but when I enter mine, it just reverts to 'Enter Boat Name'. I thought it may have been because I only registered late in leg 1, but its the same in leg 2.
Great tool.
Post by iconPost by zezo | 2009-08-10 | 20:26:46
Hi,

Maybe the special character in the boat name is treated differently by the server now - that happens with french accented characters too.
Post by iconPost by pcayrol | 2009-08-10 | 20:39:26
I think is't a problem linked to the space character...

Why the winds are not entirely correct ? Do you have a solution ?

Paco
Post by iconPost by zezo | 2009-08-10 | 21:14:42
Just got back from Greece. Will take a look (but maybe tomorrow because we finished the trip with 60Nm (12 hour) night sail from Samothrace to Kavala, followed by 5 hour drive)
Post by iconPost by TSIONA | 2009-08-10 | 21:12:52
Hi Cvetan,
Welcome home, we are all waiting for stories from real sailing.
Did you learn something new ?
Thierry
Post by iconPost by Frankfurt | 2009-08-11 | 11:51:58
Hi Cvetan,
I am glad you are back again.
I had some difficulties lately:
- In the popup window, the heading and TWA seemed to be wrong at times. Very confusing.
- And - as you know - your winds do not entirely match the winds in the game. As a consequence, the track calculated by you is neither correct nor the best one. Further consequence: You cannot win the game relying mainly on Zezo. This is frustrating (and not what you announced on your starting page).
Nevertheless, I sent you a donation. Please continue.
Frankfurt
Post by iconPost by zezo | 2009-08-11 | 20:58:00
You have a point, but it was never possible to win only by relying to the site. The optimizer is stupid and may make bad strategic decisions sometimes, the tracks are not exact even with the exact winds. For example small changes in heading are ignored - otherwise you would have to make small adjustments every half an hour. That's OK most of the time, but there are some quirks in the pro polars that make exceptions.

There is even simpler reason you can't win only by following the site - there are 1000 more people doing the same thing so you must find some other advantage :)

Thanks for the donation. Donations show me that people appreciate my work and it's worth developing the service further.
Post by iconPost by BennyLava | 2009-08-14 | 11:02:46
I also got the idea that this leg the predictions and suggested tracks were not as accurate as usually in the VORG and former legs. Hopefully we are able to go back to the old days.
Post by iconPost by Frankfurt | 2009-08-11 | 23:59:49
Thanks, Cvetan.
1. I have some other trick beyond Zezo - which I will not give away here.
2. I was very happy in the old days, when your winds corresponded with the winds in the game. Any way back to the good old times?
3. What about the discrepancies in the popup window?
Thank you for your effort.
Frankfurt
Post by iconPost by Frankfurt | 2009-08-12 | 00:01:10
Oups, sorry! I forgot I should have pressed "Reply to this post".
Post by iconPost by llamedos | 2009-08-13 | 18:10:06
Hi Cvetan,
I'm not sure were the fault is but your polars seem to be different to the games at the moment. Your chart is showing to use the Genoa at a TWA of 120 deg but the game gives a faster speed with the Light Spinnaker.
I am currently in a wind speed of 5.9 Knts at 51 deg on a heading of 299 deg at 4.2 Knts using the Light Spinnaker. If I select the Genoa it gives me a speed of 3.2 Knts and 2.1 with the Solent
Thanks for all your efforts.
Llamedos
Post by iconPost by zezo | 2009-08-13 | 21:35:26
Thanks for noting that. The figaro polars have different number of data points for the different sails, which was not the case wit VOR and TDF, so I wrote new procedure to merge polars that's probably buggy. Will take a look tomorrow.
Post by iconPost by zezo | 2009-08-14 | 12:32:45
It's not exactly a bug, but still annoying. Problem is with 4Knts polars - the Jub and Genoa have peaks at 120 and the Spi kicks in at 125. Take a look at http://zezo.org/figaro/polars/graph.pl?mode=0&minwind=0&maxwind=1&sail=4&action=Plot+all+sails

The problem is that best sail is calculated for only one data point and it becomes incorrect somewhere between 4 and 6 knots wind speed. In a perfect world those peaks would be part of the Spinnaker polars but the game has its quirks.
Post by iconPost by Belladonna | 2009-08-15 | 17:22:49
Hi Cvetan,
will you be helping us for the last leg ?
Thank you anyway.
Bernard
Post by iconPost by zezo | 2009-08-15 | 21:48:19
Yes. This time I'll be online during the start too :-)
Post by iconPost by Belladonna | 2009-08-19 | 12:42:59
Cvetan,
I did it completely wrong this time ! We have a proverb that says in french "quand les mouettes ont pied, il est temps de virer" which can be translated into english by "when the seagull are within their depth, it's time to tack !". I forgot that and went ashore twice !!! Too bad.
Post by iconPost by TSIONA | 2009-08-16 | 01:08:16
Ok !
Family holidays start tomorrow, but don't worry my laptop will be with me all the time.
This time let's make a good performance, maybe with an even better Optimizer, last leg i fell asleep one night...
I still have in mind to "eat" the Frogeater !!!
Thierry
Post by iconPost by Frankfurt | 2009-08-19 | 14:22:55
Dear Zezo,
This is an answer to the "La Solitaire Leg 4" thread. Unfortunately, I cannot give the answer in that thread.
I would like to make three points:
- I know those little details that you enlisted - and perhaps some more. What I urgently need is a correct routing optimizer. Your tool, Cvetan, was fantastic and perfect and everything I could think of during some of the VORG legs. Thanks again for that. I am longing to get back to that state.
- Today, August 19, the winds in Zezo are 20° off! I almost missed that point at 8h! (I did not, fortunately, by mere chance.)
- Nevertheless, your tool helped me yesterday to discover the north route, which brought me from 54th to 4th place over night. Thanks.
- And finally: Dear all, don't forget to send Cvetan donations. I already sent donations a couple of times.
Frankfurt
Post by iconPost by zezo | 2009-08-19 | 14:41:58
Thanks for the comments.

The GFS model was delayed by 20 minutes this morning and that caused the wind update to fail. That's the risk of early forecast updates. This happens once in a while and I should handle it better. Right now the winds used by the optimizer should be in sync, but the display still shows the previous data.

You are correct about the VORG - It was almost perfect, but there was enough time for development and debugging. That's a problem with shorter races with a lot of new features - the race is over by the time the optimizer could be made really good. The 0.25 degree is only part of the problem here. Boat polars are another - the VOR polars were smooth add well suited for the purpose, but the Figaro PRO sails ave some anomalies.
Post by iconPost by Frankfurt | 2009-08-20 | 01:03:47
Thanks, Cvetan, anyway.

I saw that the winds used for calculation seemed perfect today, as well as the 20h forcast. Thanks, that was very helpful. Only the display during the day was out of sync.

Nevertheless, Cvetan, it is due to you that I was able to finish 4th this time. I don't want to be playing VR without your tool. I just love it.

Thanks again.

Frankfurt
Post by iconPost by Belladonna | 2009-08-20 | 18:05:57
Hello Cvetan,
I've arrived now.
Thank you.
Will you be supporting the Istanbul Europa race ?
See you soon on your Big Bird,
Bernard
border
Topics list
Posts
border
1
border
border
5
border
border
Copyright 2009 by ZEZO.ORG. All Rights Reserved.