Author Message

<  Help and Support  ~  Information on WP program and Scheduler in LS Beta player

Suratna
Posted: Sun Mar 06, 2011 2:29 pm Reply with quote
Admiral III
Level 20
Joined: 02 Jul 2009 Posts: 8806 Location: Thailand TH
Hallo ALL

I see some discussions in the topic of the FR S3-21 Chiloe race on the working and differences of WayPoint (WP) scheduling and the Scheduler.

I have tried in a separate document to explain it in some more details with some screenshots from LiveSkipper Beta

https://docs.google.com/document/d/1FTD3VrqZmDAPWVDRFNsR6eciwgc3wehk0l5wUvy4dNQ/edit?hl=en_GB&authkey=CMmX_d0J


If You have any questions or suggestions to improve the document please send a post (or PM to me)

_________________
Ard-- http://vor201718.blogspot.com/ ; http://paper.li/ardsur/1345867159#
View user's profile Send private message
droopynj
Posted: Sun Mar 06, 2011 3:07 pm Reply with quote
Commodore
Level 14
Joined: 21 Nov 2010 Posts: 52
That's great explanation here. Very good job.
I used to set my WP for the scheduler mode (purple) at 10, 20, ... it might be wise to change to 11, 21, ... in case the LS server is a little bit late to avoid an early turn.
View user's profile Send private message
Suratna
Posted: Sun Mar 06, 2011 11:59 pm Reply with quote
Admiral III
Level 20
Joined: 02 Jul 2009 Posts: 8806 Location: Thailand TH
droopyNJ_Reva wrote:
That's great explanation here. Very good job.
I used to set my WP for the scheduler mode (purple) at 10, 20, ... it might be wise to change to 11, 21, ... in case the LS server is a little bit late to avoid an early turn.


that is correct. If you hit the refresh exactly at what your computer calls xx:00 it might be some seconds too early, or it is the right time but the server has not finished with all its calculations and "strange" things may appear.

I wait always till xx:01 or xx:11 before refresh, you have about 9 min, as it does not matter when you refresh as long it is well before the next 10 min update.

_________________
Ard-- http://vor201718.blogspot.com/ ; http://paper.li/ardsur/1345867159#
View user's profile Send private message
Todoncio
Posted: Mon Mar 07, 2011 2:56 am Reply with quote
Acting Sub-Lieutenant
Level 9
Joined: 06 Nov 2010 Posts: 32 Location: Austin, TX USA US
I had no idea about these timing issues. I've just launched a boat in an old race to experiment with some of these details. What I don't understand is why it is so complicated. Can't the algorithm be fixed to behave as expected?

Also, I think a great feature would be the ability to combine scheduler & waypoint mode into one track. That way we could use purple marks until we get to a gate or obstruction, then switch to green until we get clear.

_________________
Todd
The Sailor Formerly Known As "Siren"

View user's profile Send private message
pam86
Posted: Mon Mar 07, 2011 3:51 am Reply with quote
Commodore
Level 14
Joined: 28 May 2008 Posts: 5225 Location: Santiago,Chile CL
Quote:
Suratna wrote:
hallo JP

let me try to explain the WP

if e.g. it will take 9min. to reach the next Mark on the WP schedule.
your boat will sail happy for 9 min.
then it reaches the WP and waits , sit still, does not move, does not change heading , does not change sails....

Ok. This is clear. The time you talk about is the time when the boat reaches the WP, not the time associated to the mark used to program the WP.
The WAYPOINT Prog insures the POSITION where a change of sail/heading will occur,
The SCHEDULER Prog insures the DATE when a change will occur.

Even if the WP may generate some time loss, one addiionnal problem must be considered with the absolute knowledge of a future position in the Schedule mode.
Our settings as sent to the server are not exactly what we believe.
For example, while writing, Pam 86 parameters as diplayed are:
Heading -111, TWA 358
not exactly !
the server knows Pam86 as
hd 268.384
...
wsa 290.906


this must be read as
Heading= 358.384 and TWA=-110.906

So, whatever it may appear, we are never sure that our boat will be at a stricly given position at a given time.

Hence the choice to pass gates or rocks is:
- use WP prog and be sure to pass exactly close to the mark (as close as 10cm as Denosya said) or at the very pixel close to the wreck limit, while loosing between 1 to 9mn of racing time
- use Sched prog, pass a little bit further from the mark or the rock, loosing some time in the extra course and being never really sure to make a sail/heading change too early impliying a missed gate or a wreck


The choice is "cornélien" Smile Smile

Obviously, all this may be optimized by the player but at his own risk.

BUT it remains that to have a 100% guarantee to pass a gate or a rock : use the WAYPOINT prog mode.
And as all the insurances in the world: you have to pay for ... some minutes

Jean-Pierre

PDT_Aliboronz_24 PDT_Aliboronz_24 PDT_Aliboronz_24 PDT_Aliboronz_24 PDT_Aliboronz_24 PDT_Aliboronz_24

Beer please ! too much time to write this in a bad English !... sorry, my friends !
View user's profile Send private message
pam86
Posted: Mon Mar 07, 2011 4:03 am Reply with quote
Commodore
Level 14
Joined: 28 May 2008 Posts: 5225 Location: Santiago,Chile CL
droopyNJ_Reva wrote:
That's great explanation here. Very good job.
I used to set my WP for the scheduler mode (purple) at 10, 20, ... it might be wise to change to 11, 21, ... in case the LS server is a little bit late to avoid an early turn.
Denosya as guaranteed elsewhere that the "context" of a mark set at 00mn of a 10mn slice will SURELY be executed at the next position update as the context for the last passed 10mn.

When you set Waypoint or Schedule ON, a POST command sends all your marks and their context, immediately.

A reload from the browser or from the LS interface has no action on your programming.

JP
View user's profile Send private message
pam86
Posted: Mon Mar 07, 2011 4:40 am Reply with quote
Commodore
Level 14
Joined: 28 May 2008 Posts: 5225 Location: Santiago,Chile CL
Suratna wrote:
...

that is correct. If you hit the refresh exactly at what your computer calls xx:00 it might be some seconds too early, or it is the right time but the server has not finished with all its calculations and "strange" things may appear...
Yes, but a reload at 00 will only produce an erroneous display if the data received is not really the fresh one.
But this has no effect on your programs.

To have a really updated display, better to wait 1mn.
Better also to reload twice than once AND to move the map to see any weather change in the arrows.
I don't know why and when but sometimes all the data is not correctly displayed after the first reload.
Note hat I NEVER use the LS Reload. Too strange things with it Crying or Very sad Crying or Very sad

JP
View user's profile Send private message
pam86
Posted: Mon Mar 07, 2011 4:48 am Reply with quote
Commodore
Level 14
Joined: 28 May 2008 Posts: 5225 Location: Santiago,Chile CL
Siren CFORC wrote:
...

Also, I think a great feature would be the ability to combine scheduler & waypoint mode into one track. That way we could use purple marks until we get to a gate or obstruction, then switch to green until we get clear.
This has been discussed in the French forum.
Sure, it will be a nice feature BUT as the WP mode doesn't take care of time and the Scheduled mode don't take care of the position, the combination will be somewhat tricky to use.
This will perhaps be made with some additional conditions on the sequencing of actions.
Ex: "if a WP is not passed at date D of a timed mark (Sched.), don't execute the timed mark context"
and probably more conditions to avoid suprises !...

JP
View user's profile Send private message
Suratna
Posted: Mon Mar 07, 2011 8:12 am Reply with quote
Admiral III
Level 20
Joined: 02 Jul 2009 Posts: 8806 Location: Thailand TH
Hallo JP and all

I was not aware that some slight inaccuracies in heading , TWA etc are sometimes made, must be the rounding off in calculations or so.

Anyway it supports the feeling that we need some play in timing (allow some 1 min variance) and location (have some pixel maybe for the risky types amongst us 1/2 pixel between the boat and the rocks)
And be honest, mostly the time differences between the podium sailors is more than a couple of minutes Laughing

I normally use the reload button in the LS game as this works much faster, that is in the Beta.

Only when NOAA uploads have come (0,6,12,18 hts gmt) I use the refresh browser button , or sometimes even close browser, to wipe cache and restart all over. This to prevent that the LS server works with the fresh NOAA data, while the screen would see only the old NOAA data.

_________________
Ard-- http://vor201718.blogspot.com/ ; http://paper.li/ardsur/1345867159#
View user's profile Send private message
Sowslo
Posted: Mon Mar 07, 2011 9:00 am Reply with quote
Admiral
Level 17
Joined: 24 Aug 2010 Posts: 2069 GB
For those who are having difficulty comparing the classic with the beta scheduling.

In terms of programming and future outcomes the orange mark scheduler behaves the same as the classic scheduler. The only 'difference' that I can see is that in classic you type the direction or angle to wind while in beta it is set graphically. Classic you can only set at 10 minute intervals from current time while beta attempts to show a 'real time' even though that too is only activated on the next server update after that time.

Of course the big difference is that classic offers free players 10 wind / sail changes while beta offers only 4 to free players and 10 to those who pay the subscription.

(perhaps this hint to players that revenue from clicks on advertising is not covering the cost of maintaining the game is what upsets more people than the transition from classic to beta)

_________________
View user's profile Send private message
apocryphe.
Posted: Mon Mar 07, 2011 9:38 am Reply with quote
Admiral
Level 17
Joined: 13 Feb 2009 Posts: 365 EC
Hello,


Imagine !
you participate in a race during 15 days Rolling Eyes
It's very very very important if you lose 0.235468° Laughing Laughing Laughing

It seems to me, it most important to chose the best way Wink

Suratna you made a very good presentation with these examples.
Good Job.
View user's profile Send private message
Schakel
Posted: Mon Mar 07, 2011 10:03 am Reply with quote
Vice Admiral
Level 16
Joined: 15 Aug 2010 Posts: 208
loco del Rey wrote:
Hello,


Imagine !
you participate in a race during 15 days Rolling Eyes
It's very very very important if you lose 0.235468° Laughing Laughing Laughing

It seems to me, it most important to chose the best way Wink

Suratna you made a very good presentation with these examples.
Good Job.


Well, I look at it this way Wink :

I use my 10 waypoints to plan my course for the next 12 hours. On averay I loose, say, 5 minutes per waypoint, so that's a little less than 2 hours per day. On a 3-month round the world race, that amounts to about a week of waiting at waypoints.

I now know why I am currently ranked 320th in LSWR. It has nothing to do with my sailing skills. Wink Wink

Bye,
Philip
View user's profile Send private message
pam86
Posted: Mon Mar 07, 2011 9:21 pm Reply with quote
Commodore
Level 14
Joined: 28 May 2008 Posts: 5225 Location: Santiago,Chile CL
loco del Rey wrote:
Hello,


Imagine !
you participate in a race during 15 days Rolling Eyes
It's very very very important if you lose 0.235468° Laughing Laughing Laughing

It seems to me, it most important to chose the best way Wink


Sure, of very small importance, but I have in my LS archives a very precise explanation, by a very good player, about how not to loose time at a WP Smile Smile Smile Smile Times change !
Quote:

Suratna you made a very good presentation with these examples.
Good Job


Sure ! very good job that might have to be translated in Spanish and French. Thanks, Ard.

The conclusion is, whatever the prog mode used, you risk to loose 9mn59s max either stopped in WP mode or running further in a Sched. mode.

BUT ! 10mn are sometimes very much as some wins, even on very long races, have been made with a few boats in the the same 10mn.
Example, NYC-Bordeaux:
1 (8d 2h 49m 50s) loco tibenou
2 (+ 1m 11s) la guineu
3 (+ 1m 21s) mediterrados
4 (+ 6m 3s) alorsoui

JP
View user's profile Send private message
pam86
Posted: Mon Mar 07, 2011 9:39 pm Reply with quote
Commodore
Level 14
Joined: 28 May 2008 Posts: 5225 Location: Santiago,Chile CL
Sowslo wrote:
For those who are having difficulty comparing the classic with the beta scheduling.

In terms of programming and future outcomes the orange mark scheduler behaves the same as the classic scheduler. The only 'difference' that I can see is that in classic you type the direction or angle to wind while in beta it is set graphically. Classic you can only set at 10 minute intervals from current time while beta attempts to show a 'real time' even though that too is only activated on the next server update after that time.
Absolutely well said, Barbara.
The "minute" programing in BETA is misleading as the true time unit in LS is 10mn.
If the marks could be only put at 00mn of each 10mn slice, the boats will progress the same as actually and it will be less confusing for the players.
By the way, avoiding Ard to spend so much time in his brilliant explanations Smile
A lot of problems and words about something which not have to exist !

JP

About numerical /graphical programming: what is the simplest, the fastest, the more precise one ?
Have you already seen precise numerical data set via a graphical interface like in the Beta ?
About time resolution in LS: all the timestamps are defined with a resolution of 1ms (ex: "t 1299535807356" where 356 are milliseconds)
View user's profile Send private message
pam86
Posted: Mon Mar 07, 2011 10:51 pm Reply with quote
Commodore
Level 14
Joined: 28 May 2008 Posts: 5225 Location: Santiago,Chile CL
Suratna wrote:
Hallo JP and all

I was not aware that some slight inaccuracies in heading , TWA etc are sometimes made, must be the rounding off in calculations or so.

I don't think that there is any rounding in calculus. All the calculations appear to be made in floating point with various degree of precision
Examples: as sent to the server
- the accuracy of our positions is with 14 decimals,( returns with 5 from the server)
- the TWA or heading with 3 decimals (ex: angle="267.833")

The inaccuracies may occur when fractional parts are cumulated .
So, you can verify that, sometimes, a specified TWA returns from the server with a shift of 1° apart from the specified value.
And, 1° of shift may imply a shift in speed more than 0.5Knts under some circumstances.
Quote:

I normally use the reload button in the LS game as this works much faster, that is in the Beta.

Faster because it does not reload all (weather, background image, etc . But, it happens that the display is not as it would be with a browser reload.
A browser reload is more sure.

JP

All of this may be checked using Firefox/Firebug.
View user's profile Send private message

Display posts from previous:  

All times are GMT
Page 1 of 3
Goto page 1, 2, 3  Next
Post new topic

Jump to:  

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum