New Public OMP Server

[icedown]

I've set up an OMP server at icedown.net port 1515.

Available craft include:
DGXR1 at Wideawake
XR2 at Wideawake
XR5 at Wideawake
DeltaGlider IV at Wideawake
Arrow at Brighton Beach
G42-200 at Wideawake
Shuttle-A at Brighton Beach

Other craft/locations considered.
Support available at my username at gmail

Big Thanks for Ren for helping me get this up and running!

 

[Ren Dhark]

I've set up an OMP server at icedown.net port 1515.

Glad it works now.
But I'm still not able to stay connected. With STUN I get Wrp 0.1 and with custom port forwarded to my client I get disconnected after a short while. You may need more test pilots to confirm it.

DeltaGlider IV at Wideawake
Arrow at Brighton Beach

We had to remove DeltaGlider IV from OMP, you will notice terrible lag if only one pilot will choose the DeltaGlider IV on OMP!

The Arrow is ok, but only the old release included in UCGO20! The new Arrow will cause same terrible lag and if they crash, you can forget playing OMP till he leaves. The number of debris spawned at destruction has to be set to 0 in the Arrow config file (same as for DG IV).
If some guys (not only 1 or 2) would test it on OMP for a while, you could find out what to disable in the arrow-config and then – not before – add the new tested vessel to a public OMP server. ATM nobody is testing new vessels on OMP, beside you I guess.
Actually we need new pilots first! In the average I had 1 pilot per day over the last 2 years, 8 pilots online at same time was most. Post here

PM if you need any further help.
Ren

 

[Face]

BTW: if we could get some pilots online more frequently, I could start to test the new NTP mechanism (actually it would not be NTP anymore on clients, but more like IEEE 1588). And maybe that would also get me a bit more involved in OMP again :stirpot: .

 

[Ren Dhark]

OK Face, I start a new thread here where test pilots can register for a upcoming beta test. If we get enough pilots for a more frequently appearance, you can start dusting your keyboard

Meanwhile I will keep my eyes open for a PTP Hardware Clock, could be a good idea to for a master.

 

[RacerX]

count me in for testing Rhen/Face email me when you want to test. And yes The Deltaglider IV is not recommended unless Face needs to or would like this as part of the testing. Face? Also Axertan may be intrested if you would like to PM him as well. Also maybe Jango may want to contribute as well. I can also ask PDF he will most likely want to get in as well. Rhen are you retiring as hoster my friend?

 

[Axertan]

Likewise, I have some free time now and again, you can PM me if you guys would like a test. Good luck with the hosting, icedown

 

[Face]

Thanks for the feedback, guys! I might just want to test the new timestamp sync mechanism, so a limitation in used vessel classes would actually reduce side-effects.

I can't promise a specific date, but I guess I'll get some time in the next 2 month for OMP. If I've got anything new ready to test, I'll drop you PMs.

cheers,
Face

 

[PDF]

I'm In, looking forward to shooting some videos in OMP

 

[Ren Dhark]

Beta tester

Test pilots pleace register here:

http://orbiter-forum.com/showthread.php?p=427068

 

[RacerX]

well let's see we got 2 north america,czech rep.,swiss alps,austria. (I think that's right). we need people from the far east, russia, australia, south america,african contenents. This would make for an ideal testing environment would it not face? the time to test across different time zones might be a little tricky. I don't mind staying up late on a weekend if that helps

 

[Face]

well let's see we got 2 north america,czech rep.,swiss alps,austria. (I think that's right). we need people from the far east, russia, australia, south america,african contenents. This would make for an ideal testing environment would it not face?

Indeed. World-wide distribution of testers would be great.

 

[Ren Dhark]

Ren are you retiring as hoster my friend?

Negative!

I can't retire - I just started :lol: - again...

icedown's server could may act as a backup for current 0.7.1 if we start testing something new on my server. But his server seems to have still some problems - which brings us back to the topic - sorry Ice

 

[RacerX]

Negative!

I can't retire - I just started :lol: - again...

icedown's server could may act as a backup for current 0.7.1 if we start testing something new on my server. But his server seems to have still some problems - which brings us back to the topic - sorry Ice

ok rhen cool. I am glad you are
still hanging in there my friend! Awsome! Icedown I don't think I have had the pleasure of meeting you but looking forward to meeting you during testing! Anybody know how to get in touch with carmen? He was in singapore. He would be a great asset as well.


EDIT: Just tested Icedown's server everything was reachable and syncing correctly for me. I would though recommend the "Deltaglider 4-3" starting area in the server.xml to be removed. Unless of course it is now working and clients are not crashing anymore when a DG-IV (client) spawns in with a stock deltaglider already in-game. (I think that was the issue before will have to look back in the thread for confirmation)

 

[jangofett287]

I remember nailing that one down. When a DGIV is spawned that wasn't there at the start of the simulation, the simulation is automatically paused by the DGIV's module. This naturally screws your sync to hell. There might be something else causing crashes, but that was the main reason we couldn't use them.

 

[Axertan]

I remember nailing that one down. When a DGIV is spawned that wasn't there at the start of the simulation, the simulation is automatically paused by the DGIV's module. This naturally screws your sync to hell. There might be something else causing crashes, but that was the main reason we couldn't use them.

Main problem with DGIV was OMP crashing when DGIV was destroyed (original vessel gets deleted and many new vessels (debris) are created). I think it was possible to have it in the "chicken mode" so it won't disintegrate, but every client had to have it set in that mode as well. We then dropped it from the server. I *think* that was the main problem.

 

[RacerX]

hey Face with the NEW syncing does that mean this will be fixed also as a result?

 

[Face]

hey Face with the NEW syncing does that mean this will be fixed also as a result?

Not directly.

The new syncing is exchanging the distributed NTP solution for a centralized IEEE1588 solution. This means that only the server will use NTP to synchronize the internal clock to UTC, but all clients will use a much faster/sleeker and hopefully more precise clock sync mechanism, so the packet latency can be measured with more accuracy, and the RK4 compensation will return better results.

However, this will not make the approx. 400m/s speed go away. Imagine a synchronization difference of only 1ms. It would make your vessel offset by 0.4 meters. Now imagine a jitter of 10ms across the whole system... it will jitter the vessels by 4 meters.

The really big jitter in the video is from a non-synchronized client, I suppose, but it demonstrates nicely what would happen without synchronization. With synchronization, the 4 meter jumps are annoying, too, so I decided to reduce the speed issue by making earth standing still.

The real solution to the problem of rotating-earth land-jitters would be a system that transmits positions w.r.t. e.g. bases instead of w.r.t the Gbody.

regards,
Face