Woot. You all knew this was coming….Here we go with Hotlaps for Havok!!!
SL Sailing now has access to five sims of navigable water on the Beta Grid for Havok 4 testing. As discussed in the prior post, the Beta Sailing Sims include the NYC Clubhouse, the Bismarck startline, and many race buoys that SL sailors will immediately recognize. The weather is surprisingly nice for sailing on the Beta Grid, since it has relatively few people or active scripts, and the chance of getting plowed under by some foolish person in a 900-prim power boat is very small.
I think Hotlaps on the Beta Sims might be a good way to test sailing under Havok 4. Hotlaps would allow skippers to pool information and compare observations based on repeated laps with the same course in both SL and the Beta Grid. Here’s how to get going:
1. Install the Havok 4 Beta Test Grid on your computer and log in. Remember, the Beta Grid is like Las Vegas: “What happens in Beta stays in Beta.” In other words, none of the changes you make on the Beta Grid (including notecards, wind adjustments, clothing) will come back with you to the Main SL Grid.
The last time I was on Beta the sailing sims were not yet listed in the search engine, so when you log in for the first time on the Beta Grid go to SEARCH and look up “NYC Club.” That will give you a teleport to the NYC Clubhouse in Bismarck Sea.
2. I’d suggest we start with the course that MarkTwain White just devised; it looks pretty great, includes upwind, downwind, and reach legs, and covers all the sailing sims:
3. For a good test, it may make sense to sail the course on both the Main Grid and the Beta Grid. As with regular hotlaps, skippers can sail at their own convenience and as often (or as infrequently) as they wish.
4. The windsetter should be set for hotlaps, no variance. In other words, while wearing your Nantucket Yacht Club tag get close to the windsetter and say “/44 predef hotlaps“. The windsetter will reply with the following settings: dir: 5, spd: 11.0, dir+-: 0, spd+-: 0.0, rate: 1.0. Many people like some wind variance in their Hotlaps, and that’s OK too. Just say “/44 predef default“ to get the default NYC wind: dir: 5, spd: 11.0, dir+-: 15, spd+-: 3.0, rate: 1.0. Please make a notation about which wind settings you use when you post results. (Of course you can always ignore the predefs and change the wind to anything you want.)
5. Post your Hotlaps results the usual way in the Hotlaps for Havok thread on the SLSailing.org Racing Results Forum. If you are new to Hotlaps, just look at any of the other Hotlaps results listed in any of the threads and post yours the same way.
6. Since the real goal here is to test out Havok 4 sailing and compare it to Havok 1 sailing on the Main Grid, it might be particularly helpful if skippers posted their observations, comments and ideas on that thread as well. We can then see what patterns develop across the fleet as different skippers do repeated laps over the next few weeks.
So… what are you waiting for?
Get going and move beyond that milk-toast, white-bread, Second Life Main Grid you’re stuck in.
It’s time to make your life Beta.
Go ahead and put some Havok in your Hotlaps!
20 responses so far ↓
1 Owen Oyen // Jan 9, 2008 at 3:42 pm
Folks, as most know I have collected already quite a lot of numerical data for the Havok 4 upgrade in the beta sims. I have distributed it far and wide.
I have no reason to believe performance will vary sim by sim, but it would be Very Nice if this introduction of new sims could be postponed a short time. The issue is that I collected all that comparative data in Sanchon. Now Sanchon is slated to be removed. This is rather bad.
This introduces a variable that need not be introduced and forces me to re-collect all that data in new sims. If we can just postpone this . . . or incrementally do the replacement . . . then there will be time to do a compare of data already collected in Sanchon with some other sim. I tried to do this already with Hepurn, but when sidewinder ported sims for the Hepurn course he only grabbed Jaskle and Hahne, and they have non rezzable water. So I have not been able to confirm all sims are equal in the eyes of Havok. They are just CPUs and sure, Havok as part of SL is supposed to run the same in sims, but . . . the techies reading this know I am being properly paranoid.
I really need some time and a way to do this. Can someone make sure Sanchon does not disappear until this equivalence is confirmed. I am told new Havok 4 code is about to be deployed, possibly in response to feedback Espresso, Elisha, myself and others have provided. I can’t measure the certain impact of those changes numerically with confidence if new sims are tossed in as an additional variable.
All I really am asking for is a few days of co-existence of Beta Sanchon with these other sims y’all want, after the new Havok code deploys. I can get the data collected for a Yay or Nay in a few hours.
2 MarkTwain White // Jan 9, 2008 at 4:56 pm
Wow. I had no idea that Sidewinder’s choices have created an issue for Owen. Seems to me that Sidewinder would not want to lose any valuable data over this. I would suggest that Owen IM Sidwinder directly on this topic. Beyond that perhaps Elisha and Espresso will talk to Sidewinder.
While I am pleased that sims in the USS-SL are being used in the Havok 4 beta, I wouldn’t want to see any valuable work compromised or lost here. I am sure Sidewinder will feel the same way.
3 Elisha Paklena // Jan 9, 2008 at 8:02 pm
Owen, MTW, et al… I have left SWL a message regarding this issue, asking if Sanchon be left on the Beta grid under H4 for the time being, at least until enough time has passed after the next release of the H4 build for Owen to have made a qualitative study of performance. I haven’t had a reply from him yet, but I sent him the message after office hours (SWL is in the same timezone as me, so I’m assuming he was probably gone for the day) and am not really expecting one tonight.
4 Owen Oyen // Jan 9, 2008 at 8:45 pm
Elisha, you did good stuff there and this guy at LL is pretty darn responsive.
Life is good. SWL contacted me a bit ago for a very long talk about data results on stuff and what’s coming next and how. I am going to condense and distribute to all and sundry.
For now, thanks a ton for jumping on this, E. The result of it all is . . . I should have just been quiet. An interesting tidbit we maybe should all be aware of. On beta, and maybe Main, each deploy of code or restart shuffles sim assignments among boxes/cpus. SWL told me he had done some stops and restarts over the course of the work I was doing in late December and between the two of us we think the odds are pretty good the data collected was ALREADY collected from different “sims”, in the context of being sure one cpu works like another. That would likely apply to Elisha’s tests and Espresso’s too. So . . . we were already thorough without knowing it.
So life is good and . . . an advance tidbit, chances are good that TOMORROW there will be some minor vehicle mods in some new H4 code deployed to beta. He does not know what the impact will be, but tomorrow we will commence to remeasure performance and compare to pre-deployment.
All is well. The watch below may return to bunks. Elisha has the helm.
5 Owen Oyen // Jan 9, 2008 at 8:58 pm
Clarification. For tomorrow “chances are good” means if their code tweaks pass internal QA. I’ll guess that means 30-40% probable. Call it imminent rather than TOMORROW. He aims at tomorrow, but we all know about schedules.
He’s making a BLOG post when he has the deployment of new H4 code done — so we watch for it and when it shows, that says retesting can take place.
Please do not interpret this to mean that each issue we have all noted is going to be fixed. I am anticipating baby steps. Be not surprised by improvements you observe, but do not expect them.
6 Liv Leigh // Jan 10, 2008 at 2:41 am
Why the choice to include the infamous gap in the testing course? A large part of our fleet won’t be able to go through it (say: most trudeaus, the ACC, ACA, probably the Fizz) It makes me wonder if the chosen course represents common SL Sailing conditions.
7 MarkTwain White // Jan 10, 2008 at 8:29 am
The “infamous gap” (love the name, btw
) just happened to be nearby the white mark. I only noted it in my post because it was there, like the mountain. 99% of the course is open water and the 1% that isn’t can be bypassed, not sure how that makes the course “uncommon”.
The important thing is the course covers all 5 Havoc sims. If the position of the white mark is too close to Catalina for the larger boats we can shift it to the south a bit. Opinions on this are welcome.
8 Elisha Paklena // Jan 10, 2008 at 11:19 am
The other important thing is that, of those 5 sims, from what SWL noted to me, it appears that at least 3 of those are Class 4 servers, and two Class 5, or vice versa. Which means that in addition to everything else, we’re getting to see how transitions between different server architectures are handled.
9 jane fossett // Jan 10, 2008 at 11:38 am
Regarding the test course, how about this modification (USS H1 v1.1):
I have not sailed it yet, but it lets us set the shortcut (*) as an option.
I know its an ‘option’ already since no one is forced to sail through it, but… you know what I mean.
If this map is ok, I think we’d need to add an EOW warning buoy at the corner where Santa Catalina, New Georgia, and Santa Rosa meet.
10 MarkTwain White // Jan 10, 2008 at 3:11 pm
I like the modification myself. But I don’t race all the classes that will likely use it. Looking forward to other comments on the Havoc, no wait, Havok, no wait, Havoc Course.
I guess I thot there was already one of Jacqueline’s cool EOW buoys there. I will drop one there tonight.
11 jane fossett // Jan 11, 2008 at 7:24 am
Manul Rotaru and I just did a few laps of “USS H1 v1.1″ on the Main Grid using a Defender II (lite) and a Tetra 1.2 (beta).
We both liked the new course, and there appears to be adequate room. It might get congested if several boats try to get through the Santa Cat islands together, but any race fleet should be well spread out before reaching the islands turn.
Since Hollywood sim does not exist on the Beta Grid, we will need to add a marker in the extreme North West corner of Santa Catalina and have racers pass the mark on the SantaCat side. Once MTW drops that buoy, I’ll place one in the same location on the Beta Grid, and then redo the map.
12 Manul Rotaru // Jan 11, 2008 at 7:30 am
I´m also with Liv´s opinion, and the new modification of the course, the H1 v1.1, is a very useful and good decision, now all Big Boats can join the Havoc Hotlaps also.
13 MarkTwain White // Jan 11, 2008 at 9:29 pm
I have added an Edge of Sim buoy at the NW corner of Santa Catalina. Havok Course skippers should pass it on their starboard side. I suggest we remove the “gap” as an option and add east end of Catalina as a “mark” on the course.
So then, the Havok Course (or H1 v1.1 for the numerically inclined) will follow this path in Santa Catalina sim:
1. Pass white mark on your port side.
2. White mark to east end of Catalina
3. East end of Catalina to Edge of Sim buoy (NW corner of Santa Catalina).
4. Pass Edge of Sim buoy on your starboard side (only starboard pass on the course).
14 Owen Oyen // Jan 11, 2008 at 9:59 pm
Preliminary Observations in the Havok 4 Rev Late Evening Jan 11, 2008. The other observations above in this thread dated Jan 11 were NOT with this rev.
1) Overall, there is improvement from the previous Havok 4 rev, where improvement is defined as “closer to the Main Grid’s behavior”. I know the intent at LL is to become “right” and not “same as January 11, 2008 Main Grid Havok version”, but all observations that are not numerical data will likely be described in comparison to what folks are accustomed to.
2) The straight line speed on vehicles is closer to main grid speed. It is still just a little high, but speed for boats (plural, looked at ACC and Fizz) is now closer in value to the main grid
for the same sailing conditions. Also, importantly, the jerky end-of-steering turns has gone away. When a turn ended before, it ended with a jerk, too sudden, as if there was no
angular momentum. That was annoying in the previous H4 rev and they have improved it.
3) The skid, or sideways slide, experienced during a turn has been improved. I suspect those who did not see how bad it was in the previous Havok 4 version and compare only to what they have seen on the main grid will think the slide is still horrible. I am sure it has improved and I will have measurements over the next few days to quantify that. But if you’re new to it, it will seem bad.
4) There has been a serious problem introduced that I did not observe in the previous Havok 4 version. This Havok 4 version’s “violence” magnitude in response to collisions is far more extreme than what existed before. What existed on the previous Havok 4 was certainly more violent than the Main Grid, but now it is significantly worse. An ACC running aground during a skid test will be flipped over and flung into the sky. It was violently flipped over before, but not flung skyward. The reaction to boat collisions in general is far too extreme and much more so than it was. Boats are pushed aside by the slightest collision.
5) Manul Rotaru came into the Beta NYC sims while I was there and we compared qualitative notes. He had not seen the previous Havok 4 version and was therefore more critical of what he observed than I was, because I could perceive improvement. He does not think the current version is ready for hotlaps. I don’t know. I don’t think there is any harm, but also no benefit other than . . . now that I think of it . . . one. Quantity of observations. Some things and problems may be seen in conditions no one else looks at, including me. I do agree with him that this rev is not ready for anything formal and organized, race-wise. Too many things are different.
6) Mothgirl and I are good friends and I gave her the instrumentation code that I used for the ACC data collection packages that have been distributed. She stuck it in a Fizz and we both have done Beta Havok Fizz runs. My first observation from the Fizz tonight is ominous. Problems present before have not improved on this rev. There is something amiss with the sit positions (that I can’t imagine having been affected, but they are) and worse, the boat is jerking violently much too frequently for no good reason. The ACC does this too in its wave motion, but not to the extent the Fizz is doing it. Nevertheless, the Havok cause is probably the same. I think maybe the X and Y angular axes in this H-4 rev were not corrected in the same way the jerky steering was fixed around Z. Speculation on my part. I’ll just report the raw observation to sidewinder.
7) No observations taken yet of mass anomalies. In the queue. The boats are still sailing about 1 meter deeper into the water than main grid, so be prepared to be underwater often.
That’s it for now. I will post numbers over the next two days. Then a more fundamental deploy of a new Havok 4 version is due (he said) in a few weeks, hopefully addressing the observations.
15 Surfwidow Beaumont // Jan 12, 2008 at 4:04 am
i want to use the beta - not done so b4
there is no actual SLWIND in the havoc beta at NYC be nice to test that if i could
TAG Problems
~~~~~~~~~~~~
when i logon to the beta i revert back to the first tag i had when i logged on grrr and you cant join the NYC group in Beta either
and i cant join a group i am already a member of in Havoc Beta nor can i leave it
the groups list is blank and i am stuck with the SLSF tag that i cant remove in beta
any ideas
:\
16 Surfwidow Beaumont // Jan 12, 2008 at 4:07 am
PS IM me inworld if you have a solution too
17 Surfwidow Beaumont // Jan 12, 2008 at 4:11 am
panic over i logged in and out a few times and its rectified
18 jane fossett // Jan 12, 2008 at 8:20 am
I just started a new thread at ORG to discuss Havok 4 on the Beta Grid. The forum has an easier format for posting and comments as we go along: http://slsailing.org/forum/viewtopic.php?p=6365#6365
19 Manul Rotaru // Jan 12, 2008 at 1:33 pm
Owen wrote:
“5) Manul Rotaru came into the Beta NYC sims while I was there and we compared qualitative notes. He had not seen the previous Havok 4 version and was therefore more critical of what he observed than I was, because I could perceive improvement. He does not think the current version is ready for hotlaps. I don’t know.”
A bit of a misunderstanding, but I was on only shortly, was somehow busy…
First I was very critical, yes, but after I have realized that I have the same ACC upwind speed like I have in the main grid, I said that I plan to do hotlaps there this days…
I would have already done several hotlaps there, but I still have the same issue Surfwidow wrote about, I still don´t have my NYC skipper group tag in Beta which I need for the startline. I go one trying that I can start with the Hotlaps in the Beta Grid soon. :))
20 jane fossett // Jan 13, 2008 at 8:51 am
Thanks to MTW for adding the additional EOW marker buoy. I’ve placed one on the Beta Grid as well. Here’s the new chart:
Leave a Comment