Control Keys vs Chat for Sail Trim

I think healthy discussion of how the general Tako sailing group would like to see the Tako derived sailboats evolve is a "Good Thing[tm]."

As such, from the Sailing Tips thread -- I thought a good place to start might be Control Keys, Chat, HUD, no HUD, visual queues vs numbers... Basically, the core control and feedback mechanisms -- since these directly effect "how" we sail in SL.


To start us off, I'm interested in how you currently sail? Please share through the poll, and if you like expand on the reasons why you sail the way you do now, and what if anything would influence you to move to/from control keys to/from chat & gesture.

I guess this is better place for the controls discussion than the racing tips thread.

There was no choice in the poll, but i use the control keys almost exclusively .. with +/-5 and +/-1 on the button bar. No gestures.

If i could get predictable results with control keys - in a way like the progress has been on Static's experimental boat - I could drop the controls from the button bar alltogether too.

My only point about gestures was .. that the way they are now - they make sailing a competition of user interface programmin/customization - not sailing. Because everyone doesn't stand on equal ground. Newbie with no knowledge or skills in gesture editing will always lose against the person who has their gestures tweaked for best sailing. And the way the gestures can be customized makes trims possible with "one button push" to specific angles.

I would like to see in some future boat a setting, that would be predictable and easy to use - AND would put every sailor to the same place regards to handling the boat, and not give advantage to those who extensively like to tweak their controls.

Suzanne,

I think we may have a slight misunderstanding here.

I don't customize gestures for a course. I have a stock set of gestures that I use for every course, the mainstay of which is simply a 5 degree and 1 degree sheet command. I think this setup allows me to be flexible and respond to the dynamic situation that is a race. If I had gestures set up for each bouy... what would happen if I rounded the bouy next to another racer and had to give right of way? I'd be in trouble if I was dependent on pre-set gestures to set the sail instead of being responsive to changing situations.

In effect, I am using gestures to do the same thing that you do with the button bar.... but instead of pushing the buttons with a mouse on the screen, I push the buttons on my computer IRL.

So, IMHO, based on what you are telling me, we sail very similarally, if not with the same tools. You let out 20 degrees of sheet with 4 mouse clicks... I let out 20 degrees of sheet with 4 keystrokes.

I have no concerns about sailing against folks who use preset sheet gestures for each bouy of a course... I don't think they will be keeping up with me... or with you, Suzanne. Same is true for auto-trim scripts. I just don't think the current state of the technology is a substitute for knowing how to sail and having a system that allows you to concentrate on sailing, not on the interface needed to sail.



I agree Suzanne.

I am perfectly willing to share my gesture library with any other sailor who asks. I have shared it with many sailors who have, and had discussions with many others to help them develop their own library. I often speak with other sailors about gestures for trimming, and IMHO they are no more (or less!) esoteric than programming and using the button bar that comes with the TAKO HUD.

Like I said, my gesture library is the same regardless of where I sail. I don't customize per race. So if you are interested in trying it out, let me know in world, and I'll gladly share.

I really don't think we are in disagreement, Suzanne... I think we've just latched onto a different set of tools.

Bear in mind, the gesture method of sheet trim was developed before the Tako 3.0. The button bar was released with the Tako 3.0. Therefore, gestures were the only way to precisely and consistantly manage the sheet when I started sailing. Having started with the 3.0, you had the choice of a different toolset to interface with the TAKO than I did when I started. I think our apparent disagreement probably has more to do with that difference than with any fundamental difference between our attitudes on sailing the Tako.

See you on the water!

--Faykin Odets

I never use the button bar. I use one gesture for getting into close hauled trim, and +/-1, +/-2. I never liked the button bar because I want to see as much as possible. I'm using the hud light, too.

In most of the sailboats I have raced in there are different apporach to the layout of the lines and pulleys with the aim of optimizing the RL interface to the ropes. I recall many interesting discussions about what was the best way to layout an RL boat's controls.. For instance, even in a pretty strict one design boat (the Flying Scot) people used to go back and forth about whether it was better to have the mainsheet lead from the end of the boom or off the center (both were allowed). The difference is that in SL, you may not be able to inspect the other person's customizations like you can on an RL boat.

Removing all the gesture-based control in SL will simply move some of the user interface customization outside of SL and into gaming keyboards with programmable "G" keys that allow for placing commands onto another set of function keys. I saw one of those that looked interesting at the store a couple weeks ago. If gestures are available, it is at least possible to share them in SL. I don't think we can or should get rid of user interface cusomizations completely, and except for the super strict one-design RL boats, cusomization is a reality of RL sailing. Let's just try to get rid of auto-trim scripts, and put effort into better windsetters.

LOL...

Well then... I might be able to give you an edge up by taking away your chat commands then.

I assume the reason why you use the chat commands with this setup then, is because it gives you predictable trim amounts -- not because of the other advantages of "single key to base trim" a leg, but rather the predictability of it.

The wacky randomness of of the past where, depending on current server lag, you'd get anywhere from 1 to 15 degrees (or more) for the same length of key stroke -- are a thing of the past.

This currently requires hard coded changes, but the control keys can be set so that a single tap does >exactly< 5 degrees, and a shift control key could do exactly 1 degree. And it'll do it suffering less lag then a Gesture => Chat command. If your already doing 27 degrees by counting out five F12s followed by two F11s, the same could be done using the control keys now.

HUD updates whether they're from Gesture => Chat commands, or from control keys are the same either way.

The added benefit though, of using a control key event instead of a chat command, is that the control key doesn't have to go through chat filtering to be routed to your boat, nor does it have to use relatively slow string parsing to interpret that command. Control Key events are therefore a little bit less laggy.

The button bar uses chat to talk to the boats, so it falls in along the same lines as the gestures -- although ever so slightly slower, due to having to interpret the mouse event. Probably not a measurable difference.

<laughs> Well, Static, if you get that implimented and debugged, then yes, I will have nothing to learn except the new key postion. So... please, make it real, I won't have to change my style at all!

--Faykin Odets

I use chat gestures exclusively, save for steering. I was originally against making the spinnaker easier to control, but after the advent of Suzanne's retail race timer I decided 'why the heck not?' and commenced development of the t/5 as a strictly-manual one-design alternative to the Tako 3.2. I maintain that for any two skippers of the same technical skill level, the one with more efficiently automated trim control will always come out ahead, so as long as chat commands are part of the standard Tako control system, and as long as the start lines support personal race timers, I'll probably keep using them to stay competitive.

Right now I have three grains of trim mapped to my function keys, plus centerboard and spinnaker toggles. My coarsest controls let me snap my mainsheet to 20, 30, 45, 60, or 70 with a single button press, to be modified by medium-grain +/-5 and fine-grain +/-1 trim adjustments. With the Tako 3.2 I've slaved spinnaker trim alongside the mainsheet controls such that my 45, 60, and 70 mainsheet coarse controls also snap the spinnaker to 0, 30, and 50, respectively. Medium-grain trim controls also adjust my spinnaker angle +/-10 and fine-grain controls slave spinnaker adjustments by +/-2. This doesn't provide truly automated trim, but provided that I immediately follow every raising of my spinnaker with a coarse-grain 45/60/70 mainsheet snap, I ensure that my spinnaker follows in synch with mainsheet trim.

I was very good with the old manual spinnaker under the Tako 3.1, but this new system has taken me about a month to become proficient at controlling. In the end, I think I'm able to keep my trim slightly more optimal under the new system than I did before, but it's still a bit of a juggling act to get the keystroke sequences correct; if I just once get the steps out of sequence or do the math wrong in my head, I'm scrambling to catch up with those who didn't make the same mistakes by keeping their controls fully manual.

I can't stand HUD controls, they slow my response times way down, so I stick with function keys and the HUDlite, moved off the top of my screen enough that I only see the wind angle, sheet angle, spinnaker angle, spinnaker status, and centerboard status as bare lines of text. My chat gestures aren't the most efficient, I realise, but the learning curve each time I've changed them (thrice) has been steep enough that it slows my performance way down for about a month until I learn the new controls somewhat reflexively again; a big disincentive to non-essential changes. Ideally, someone with centerboard and spinnaker status slaved to key trim points for 35, 90, and 150 degree wind angles could probably gain an edge my current system can't match. Because the spinnaker and centerboard chat controls are true toggles, though, I doubt such a system could ever be implemented effectively with on a stock Tako 3.2; the penalty for losing control/status synchronisation is too high.

I will admit that, despite my appreciation for the rush of fully-manual boathandling, chat gestures do make pleasure sailing much more stress-free and also go great distances toward ameliorating the frustration of sim lag/control synch issues. On a tack, I can tap two to four keys in sequence and be fairly confident that my trim will eventually dial into the neighborhood of my intended target once the command buffer processes. I've grown to appreciate this 'laid back' style of sailing enough that I've chosen to develop a fully automatic mode for t/5 as an alternative to its core strictly-manual control scheme, with obvious visual distinctions between boats of either mode, allowing for different classes of races.

Faykin, my post wasn't personally directed at you, and the points about controls optimized per race course etc. were just general observations, nothing to do with you.

But I think Myrrh's post puts it all better in words that I could - how far can one go with automatized gesture controls that allow precise angles for *both* mainsail and spinnaker at one single keypress. And I will stand behind my opinion - that competitive sailing has become a game about user interface customization and control tweaking - rather than sailing skills.

For my 2 cents, I would have no problem lossing the chat commands and making the Tako only control key operated. I feel that this would be closer to RL sailing and take away the advatnged gained by clever gsetures or worse - scripts.

I wholly agree with Suzanne. If anyone hasn't yet read through the t/5 specification, I encourage you do to so and contribute any comments you may have - it was created explicitly to remedy this situation. Its plans have grown a bit more ambitious since the original seed of an idea, but it should prove a very exciting one-design class for experienced SL sailors come its autumn debut.

Regarding the Flying Tako, I'm afraid the genie's already out of the bottle. For my part, I do everything I can to share my experiences with other sailors, regarding synchronised chat gestures and key 35/90/150 points of sail, both. My own gestures are always freely available for the asking to everyone, if I haven't given you a copy already. I'm perpetually curious to see what other people come up with as I'm sure there are far more streamlined implementations than my own.

It's worth noting that Static has conducted extensive testing with a full-blown scripted autotrim HUD and determined that, at least under current system bottlenecks, it can't keep up with an expert skipper using manual controls. Somewhere along the line there's a tipping point between the efficiencies of automatic and manual systems, and I suspect that my own gesture set is pretty close to that limit - my performance under the slaved trim gestures has been marginally better at best, possibly unchanged.

Mostly it's just a lot less frustrating when the the sim decides to fight against
me. Having sailed with chat spinnaker commands for a month now, I must admit that I've come to appreciate the opposing viewpoint to my own - that trim gestures allow me to sail in certain environments which would otherwise be an exercise in frustration, be it laggy sims, cluttered waterways, or just recreational sightseeing.

You'll note in the next t/5 development update, to be posted after the boat show, that I'm adding a fully automatic mode to its default manual operation. I'm not convinced that there's need for an inbetween mode - we already have the Tako 3 for that - but the t/5 will also support point-and-go recreational/tactical sailing with an enforced change of appearance and boat name, for immediate visual and start line recognition during races.

Just as with Suzanne's race timer (yes, that too was a controversial idea about four months ago), I initially set up my Tako 3.2 spinnaker gestures to prove a point about their unfairness. In practice, they've shown to be far less a magic pill than I expected - my performances within the fleet under race conditions have not changed noticably. I recommend that everyone give both styles a try, you may be as surprised as I was.

I still maintain that chat gestures give an edge amongst skippers of equal ability, other skippers maintain that gestures can never match full-manual performance. That's mostly conjecture and gut feeling on my part, though - I've raced both ways, for extensive timespans, and the results have been inconclusive. The fleet leaders have consistently been a mix of sailing styles, so maybe in the end that speaks to the equity of the Flying Tako 3's roll-your-own control schemes.

An addendum to the note, is that I've never actually used it under "real" race conditions -- only during friendly races, which have generally been only 2 to 5 racers.

A fully automated system lacks a few things that fully manual control allow for:

1. Updates occur only as often as HUD updates are sent -- if as a human skipper you wait for the HUD to update, to set your trim, then the automated system will win -- it'll do it faster, and more accurately then you can. But if you know, in general terms, that after rounding marker X your going to need to ease the mainsheet about 30 degrees, and you've got accurate predictable trim controls, you can go ahead and hit that +5 trim key 6 times then wait for the HUD to update for fine grained adjustements.

2. Sometimes optimal trim, isn't optimal I have on occasion needed to tighten my main to avoid (or reduce) being tangled with someone elses boat or a bouy.

3. No matter how optimal your sail trim is at 170 deg wind direction, your going to go slower then someone at 140 deg win direction. Now whether or not the extra distance your going to travel at 140 deg, will be offset by that faster speed -- thats for us humans to decide.

Automated sail trim would, IMHO, make sailing much more tactical -- in fact, that's all it would be about, is tactics. Picking the best line, knowing when to tack -- and if your so inclined, when to use ROW.


But... getting a little off topic... I'll be in-world most of today (5/24), after the grid comes back up. Anyone that has some free time and would be interested in experimenting with different control settings -- please shoot me an IM.

I chose the first option, that I use control keys exclusively. I probably should have chosen the second option, because I do have sheet +/- 1 and sheet +/- 10 programmed, though I use them so rarely these days that I would never miss them if they were gone. I do not use the button HUD, but rather HUDlite and the function keys. It's important to note that very often I am holding the control keys down when a large trim adjustment is needed.

I use the control keys because, while I know chat commands can give me an edge against equal opponents in a race or make life easier in a lag situation, the control keys give me the feeling that I'm in a RL sailing situation. It takes a certain amount of time to get an exact trim this way, makes the spinnaker a more difficult device to use.

It's a personal choice that I know hurts me in racing. I like the competition, but I like the sailing experience even more. I work at getting a feel for how my control keys react to my input, which is how I manage to do decently at all in the races.

However, many people race to win. I don't deny chat commands are vital for that. And Myrrh brings another good point that, for just simple leisure sailing, chat commands bring an easy automation to the system.

Control keys have served me well until the new Channel Island sims that surround Hollywood. If it's just a couple people sailing around, I think I'm okay. But the load of ten sailboats on these low-load sims have proven too much a strain for the Tako to handle as it's scripted now. I think it is great that the scripts are being looked at again to reduce the weight, which I'm sure it will not only improve performance in the Channel Island sims, but in other places I sail as well.

As for my experiences with the beta Tako... the first time I sailed one of these (maybe two weeks ago?), I found it a great improvement in performance. The control key commands had become predictable for each tap or hold down, but there was still a period of fine tuning the trim such that I felt it was similar enough to the Tako 3 not to be a problem for my sailing habits.

This feeling completely changed with the lastest beta Tako. I felt the control key commands had become twitchy, that I had to fight the system to get it to do what I want rather than work with it. A steady tap for mainsail trim is too slow. Trying to tap for a larger adjustment requires figuring out the rhythm and timing to the taps to get it to adjust properly. Get that wrong, you can find the sheet angle shoot way off the mark that you intended. The spinnaker adjustments are predictable but far too slow to get a large angle. Most of all, forget about holding down the key to get a medium or large sheet adjustment.

As the beta Tako is now, it encourages the use of chat commands. If you need a huge adjustment on the spinnaker, you might as well do the chat command because to use the control keys will take you forever. If you want to adjust the sheet angle by a considerable amount, it's to your advantage to use the chat command because trying to time the control key taps is twitchy. The control keys then become best used for only fine tuned adjustments, like the sheet +/- 1 command that people so often use. I hate it. It goes completely against my perferred method of sailing.

However, I know that the overwhelming number of sailors use chat commands in a race because they find the control keys too erratic. I wonder, though, if the control keys were made predictable, if people would bother switching to the control keys. I suspect not -- that they'll continue to use the chat commands save for the fine-tuned adjustments from a predictable control key.

I like the Flying Tako controls as they are now. If there were a way to decrease the script load without changing how the Flying Tako reacts to the control keys or chat commands, then that would be the route I wish the scripters would take. It worries me that this is not so, but I did say that I could relearn how to use the Tako. But would I like it as much as I do now?

Just a note -- each time I crack out the betas, there is often subtle changes to the control code from the last time. This most recent attempt in hollywood was with an exponential increase in sail trim -- if you held the mainsheet key, or tapped very very quickly. This didn't work out too well...