Daily Archives: October 17, 2011

Nemo 2 Interview

Last year Nomad Zamani and Glida Pilote got together and crafted Nemo, a six meter keel-boat inspired by the Laser and Flying Fifteen. Well, Nomad and Glida are at it again; they recently launched Nemo 2, a truly innovativeย  pocket-racer.

Nemo’s creators kindly agreed to talk about their new build on MetaverseSailing, and their interview is posted below!


JFos: Hi; thank you for taking time to talk about your new boat!
Perhaps we can review some basic nuts-and-bolts first, since many readers may not be familiar with Nemo.
Can you tell us why you decided to build the original Nemo? After all, there were several other dinghy racers available at that time, and there are even more now. What was your goal when you started this project?

Nemo: We wanted to make a boat that would encourage new people to try sailing in SL, particularly beginners. There were already free boats available, but we felt that they all had aspects that reduced their appeal somewhat (time limited, quirky styling, or whatever). So we made a sleek, good-looking boat that was fun to sail and easy to use, and that didn’t delete itself after a certain time. Now, in mid October 2011, distribution is fast-approaching an amazing 10,000 free Nemos.

JFos: I knew the first version of Nemo was popular; in fact I’m still using a Nemo Race pic as the banner for this website. ๐Ÿ™‚

Nemo: I’ve always liked that picture – it says “shared fun on the water,” which is what it’s all about.

JFos: Wow, 10,000 free Nemos, I guess I never really appreciated how popular the boat turned out! That actually raises my next question, why did you decide to upgrade that boat to Nemo 2?

Nemo: Nemo 1 was based on the early Tako sailing engine, and an update was always on the cards, but we wanted to launch the boat to gauge interest. The response was good, so work on Nemo 2 started within a few weeks, and took about a year and a half. The main aim with Nemo 2 was better realism. There was no particular reason to release Nemo 2 now, other than the feeling that it was time to do it – that a significant milestone had been reached. There are still some ideas to come, and we’re planning to do some incremental releases to phase them in.

JFos: The original Nemo came in two flavors: a basic version for free, and an upgrade version that sailors could purchase for a small charge. Have you continued that arrangement for Nemo 2?

Nemo: No, there is now only one version. The reason for the free and retail versions of Nemo 1 was to give us a way to estimate the number of ‘conversions’ – people who tried the free boat and took enough interest to buy the low cost retail version. For Nemo 2, we reconsidered. Since the primary purpose of the Nemo is still to attract new people to sailing, and because most newbies don’t have any money, we elected to drop the retail aspect – it’s more encouraging to give them full access to all of the boat’s features right away. It also means that there is only one version of the boat to produce, so development effort can be focused in more productive areas.

JFos: Nemo is free? ๐Ÿ™‚ Pardon me, but lets have a 30 second break here, so sailors can stand up and applaud Nomad and Glida. ๐Ÿ™‚
I think we know that sailors don’t go to heaven, they’re too crusty and foul-mouthed for that… but if Nemo’s free, maybe you two deserve a special pass. ๐Ÿ™‚
You’ll have to stand in line behind the builders of the free Shelly, BBK, and Fizz that came earlier, but wow that’s not bad company. ๐Ÿ™‚ Nice work!

Nemo vendors

JFos: OK, lets continue… Where can sailors get Nemo 2? I hear that estate owners can get Nemo vendors; what’s the arrangement for that?

Nemo: Nemo 2 and its vendors are all free. The official Nemo display (Nemo HQ) is at Nantucket Yacht Club, but anyone can get a boat from any Nemo vendor, or from any Nemo (the boats themselves also function as vendors). Estate owners are more than welcome to rez a vendor at their dock, their training area, in their shopping mall, etc.

JFos: That should make them very popular!

JFos: Let’s now talk about a few Nemo 2 details, ok? How has Nemo 2’s appearance changed? Is it the same boat, or new features?

Nemo: The Nemo is still a very pretty 6m keelboat with a single sail. The basic shapes have not changed, but it does have new sails with distinctive radial stripes in a selection of bold colours, and the crew now lean in and out with heel. Other changes are more subtle, like the wind type indicator around the base of the mast, and the round trim indicator in the top half of the bulkhead instrument. We are currently working on sculpted sail shaping which will be in a forthcoming update.

JFos: Are there changes to the underlying build? I understand the new boat is now ‘Mod,’ and prims are user adjustable.

Nemo: While the basic construction is the same, Nemo 2 combines the two variants of Nemo 1, with extra flexibility into the bargain. It now has both race wind wind and a simpler (but much more sophisticated) ‘rez and sail’ type of wind, and has modify permissions. We’ve added a Settings notecard for defining custom preferences, the sail textures are full perms, and the more enterprising customizers could even replace the crew animations if they wanted.

JFos: Any thoughts about Mesh in your future? ๐Ÿ™‚

Nemo: We’re aware of Mesh, but it’s very early days for that technology, and many current viewers are not compatible with it. It does have potential, however.

JFos: I suspect you are right. I can add here that my first tests of the Nemo 2 production release had a ‘glitch’ when I rezzed it under Firestorm. In case others experience this, the problem wasn’t the boat, it was a Server / Firestorm issue, and it spontaneously fixed after two days. Thank you to Nomad for helping me trouble shoot the problem; the Nemo crew get an ‘A’ for tech support.

JFos: OK, let’s talk Nemo 2 performance issues. ๐Ÿ™‚ Is the script load the same? What’s changed?

Nemo: It’s hard to gauge script load in a boatย  that has multiple scripts, or to compare with an earlier boat that has different characteristics. When Nemo 1 was made, certain commands weren’t available, which influenced the scripting structure. With new commands available to Nemo 2, there are fewer scripts doing more work, and some very small new scripts appeared (like the collision sound handlers). Script optimizations also help things to run more efficiently. A key aspect of Nemo is that it has always been geared towards low lag – it is not controlled using chat listeners or gestures, and there is no external hud. Suffice to say, Nemo 2 feels very responsive.

JFos: I chatted with Armano Xaris yesterday. He’s a skilled racer and highly analytical sailor. He gave me a head’s-up about the Nemo 2 wind system, and Armano thought it was pretty great. Can you tell us what he was talking about? What did you do to the wind? ๐Ÿ™‚

Nemo: Oh, not much… …apart from completely reinvent it from the ground up. ๐Ÿ™‚

JFos: This sounds like a big topic. Does Nemo 2 have Wind Shadow?

Nemo: Yes. The new wind system began as an experiment in implementing wind shadow that was ultra low lag. I wanted to avoid the exponential chat lag that many boats suffer from, and to be free from restrictions like chat range and sensor limits. So I networked the wind shadow. In essence, the boats to talk to a separate server rather than each other, which adds up to a dramatic reduction in the load on the SL systems, and the scope to build a far more sophisticated shadow model.

JFos: Tell us about ‘Global’ wind settings.

Nemo: There are no settings – you simply rez and sail. The boat can still use race wind (and now restore it after a crash), but global wind is a different thing entirely. The server extracts wind data from a real world weather buoy, and uses this to create a wind profile – sequences of gusts and shifts – that is made available to all boats in real time. The profile has random elements, but is consistent for all boats. The chance of the pattern of gusts and shifts repeating is effectively zero. Nobody, including me, can predict when the next puff is going to happen, or how strong it will be.

JFos: What does networked wind do that current methods don’t?

Nemo: A networked wind system can be developed in the background without having to revise the boats, and it has scope for introducing features that are impractical in implementations that are based solely on in-world scripts.
The wind profile can be refined, and other subtle wind/boat effects can be added. The first step has been taken to model local variations due to terrain – we are currently experimenting with land shadow. This is not a preconfigured system like that used in the WWC wind setters, it is real time. The lee side of an island or building moves around with the current wind. Even when the wind veers slightly during a gust, there will be subtle changes in the shape and position of the land shadow.
The client/server architecture also means that other client types can be introduced – they don’t have to be boats. There is a wind sock at NYC that reacts to the current global wind, and we have a prototype live race map that can show the positions of boats registered for a race as they sail around the course.

JFos: The first Nemo used a modified Apparent Wind calculation; Wildwind boats do a similar same thing. I appreciate this is a five year-long discussion that dates back to issues with the Tako, but its an important performance issue. Is Nemo 2’s ‘Apparent Wind Adjustment’ the same as the original Nemo?

Nemo: The scaling to compensate for Tako speeds has gone – Nemo 2 has apparent wind done right. For example, I recently experienced some sudden lulls in global wind. For a few moments, the boat’s momentum carried it forward at a higher speed relative to the now-lighter wind. The apparent wind suddenly moved forward quite a lot, and gradually moved back as the boat slowed in line with the lighter wind. This is exactly what should happen. I didn’t think about this when I programmed the app wind code – all I did was crunch the vectors and trigonometry. When I first saw it, I had one of those moments where you go, “now, why did that…. of course!”

JFos: Have you changed the wind engine polar in Nemo 2?

Nemo: The sailing engine has been rewritten from scratch. There isn’t really a polar as such, but a mathematical function that defines the boat’s characteristic power transfer curve for all points of sail. From this, the boat’s actual speed is influenced by the size of the sail, the amount of heel, sail trim, etc. A less obvious factor is wind speed – Nemo 2 has windage and will drift to leeward, so even though a stronger wind can make it go faster, the effect is non-linear – leeward drift aids you on a run, but works against you on a beat.
So, the boat does have a polar, but it is by no means ‘programmed in’ in a simple way. The entire engine is mathematical (there are no lookup tables), and while the dynamic factors have been modeled with a certain amount of independence from each other, they do interact. For instance, heel affects speed, but not directly – greater heel increases water drag, and drag affects speed. It’s nearly impossible to predict the behavior of a system built like this, so the development of the engine was done empirically – the dynamics were adjusted and tuned over many weeks to home in on a boat that felt right overall. I don’t actually know what the polars are, and it would be fascinating to see the curves for different non-varying wind speeds (and especially a comparison of the two sail sizes as the wind gets stronger).

JFos: Sailing in SL is sometimes wondrous, but frankly it often falls short of our expectations. What do you hope will change this coming year to improve sailing on SL’s grid?

Nemo: I bet you say that to all the boat builders, and I bet they all say: ‘better sim crossings’. ๐Ÿ™‚
I’d like to see the principles behind global wind gain traction because I think the wind setter model is becoming dated. Anyone that has sailed in RL knows that nobody controls the wind – that race courses are chosen to suit the conditions rather than picking wind to suit the courses. The problem with the latter (aside from being rather bizarre to a RL sailor) is that things can get very samey – when you sail the same courses in the same wind, week in, week out, you end up being able to predict your tacks and gybes to the Nth degree. You can even use navigation tools to set waypoints for the turns you will make next week, or next month. Real sailing is not predictable and repetitive, it is a dynamic activity where things change constantly – the notion that you can practice a course that you will be racing in a week’s time simply doesn’t exist.
If SL wind had been done properly from the start, there would never have been a need to invent wind setters. The wind would have been accepted for what it is, Hay Ah’s excellent new rotating start line would have been invented long before now, and it would be natural for every sailor to know the rules of the road because every boat would be sailing in the same wind. I honestly feel that, while the wind setter model was necessary in its day, the legacy wind technology in SL is heading towards a dead end. It lacks scope for better realism, dynamism and expansion. I doubt that LL will ever fix SL wind, so my view is that, if it’s our world, our imagination, then it’s up to us to take the technology forward.
Nemo 2 and its Windmaster networked wind system is a major step towards this. By demonstrating that the technology really can be built, it will hopefully inspire others to try similar systems. Indeed, I have already had some interesting conversations with others who see the possibilities that this approach offers and want to learn more.
Over the next year, I would like to see other developers investigate networked systems with a view to eventually building something that can be used by all boats in SL. I don’t think we’ll see global wind for all within a year (it is not a trivial undertaking), but learning the systems and techniques, and possibly building the foundations of such a system, could be achieved. For Nemo, the focus will be to develop the boat further and continue enhancing the Windmaster system – and maybe do some sailing that isn’t just thinly disguised testing!

JFos: Well, those are tall orders, but they give sailors great goals to set and work for. I certainly hope they get get realized too.
Thank you for sharing your ideas, and thank you for Nemo. ๐Ÿ™‚
I learned a lot today. ๐Ÿ™‚


Addendum (JFos): There are several free boats available to new sailors in SL. Although they are free, please never think they are low-tech or ‘dumbed-down’ compared to other SL sailcraft. The Flying Fizz, the Shelly, and the BBK are all great boats with innovative, pretty wonderful scripting. It looks like Nemo 2 now joins that select group.

There’s a common saying: “Money can’t buy you love.” Well, it looks like Nomad and Glida proved it again. Nemo 2 is indeed a boat to love… but it comes free,ย  so it’s not so simple. You’ll need to sail with Nemo, earn the affection, and build a relationship.
My guess is that could take a full two or three minutes, depending on which of Hay Ah’s racelines is doing the countdown.

By then you’ll be hopelessly in love with this little boat. ๐Ÿ™‚