Page 4 of 4
Posted: Thu Jun 18, 2009 11:09 am
by DaddyHoggy
So that answers one important question and asks about a hundred more!
So sub-entities are only a problem if they're NOT flashers - OK, so what's the difference between a flasher and a normal subentity other than flashers aren't really "there" being basically point sources with no mesh...
Posted: Thu Jun 18, 2009 11:17 am
by dalek501
Nemoricus wrote:Which type was it? Was it the small satellite or the large fly-through station? It's important because if it was the former, it means nothing. It's the latter that you need to test.
Ah, only the small one. I'll see if I can find a big one.
Posted: Thu Jun 18, 2009 11:19 am
by DaddyHoggy
I retract my previous posting until new evidence presents itself!!
Posted: Thu Jun 18, 2009 11:21 am
by dalek501
OK found one I could fly through, and all was fine. No problems.
Posted: Thu Jun 18, 2009 11:24 am
by Nemoricus
Ah, excellent.
I think that flashers may not have any collision enabled, which is why they don't cause problems. If collision is disabled for all sub-entities it may solve the problem, at least temporarily.
Posted: Thu Jun 18, 2009 12:46 pm
by Cmdr James
DaddyHoggy wrote:So that answers one important question and asks about a hundred more!
So sub-entities are only a problem if they're NOT flashers - OK, so what's the difference between a flasher and a normal subentity other than flashers aren't really "there" being basically point sources with no mesh...
Flashers are handled differently, as are exhaust plumes, even though both are subentities. I dont think its easy to list all the ways that they differ, and I am not sure it will help as what is likely to be the problem is that the code is just different, and there is some kind of case that isnt handled properly in one or more of the blocks. You probably wont be able to tell by thinking about the functionality, but only by inspecing the code.
What would be good would be to get a simple test case that shows this crash (an OXP and clear instructions, I guess) and then one of the devs, maybe me, can run it in debug and see where its failing.
In fact, if its a collision problem, the perhaps enabling bounding display might be interesting?
Posted: Thu Jun 18, 2009 12:53 pm
by Screet
Cmdr James wrote:What would be good would be to get a simple test case that shows this crash (an OXP and clear instructions, I guess) and then one of the devs, maybe me, can run it in debug and see where its failing.
Fly a neocaduceus into a thargoid hive or fuel station
With the hive I need to disable all subentities, not only the turrets.
I've once had a look at the code, but I was very tired...didn't find the cause
Screet
Posted: Thu Jun 18, 2009 6:13 pm
by ClymAngus
Screet wrote:Cmdr James wrote:What would be good would be to get a simple test case that shows this crash (an OXP and clear instructions, I guess) and then one of the devs, maybe me, can run it in debug and see where its failing.
Fly a neocaduceus into a thargoid hive or fuel station
With the hive I need to disable all subentities, not only the turrets.
I've once had a look at the code, but I was very tired...didn't find the cause
Screet
Then try the same thing with a merlin. That has sub entities and it's much smaller. I would be interested to see if it goes BANG! If it does we have a bigger problem if not then it's probably a size thing.
Posted: Thu Jun 18, 2009 6:15 pm
by Nemoricus
I don't think Merlins have subentities. One would have to be added in order for that to be a valid test.
Posted: Thu Jun 18, 2009 6:19 pm
by ClymAngus
Nemoricus wrote:I don't think Merlins have subentities. One would have to be added in order for that to be a valid test.
I thought they did? Crack open the oxp and take a look. Or maybe go with an imperial courier?
Posted: Thu Jun 18, 2009 6:43 pm
by Screet
ClymAngus wrote:I thought they did? Crack open the oxp and take a look. Or maybe go with an imperial courier?
It has:
Code: Select all
<key>subentities</key>
<array>
<string>l-mexh 0.0 0.0 0.0 1 0 0 0 </string>
<string>r-mexh 0.0 0.0 0.0 1 0 0 0 </string>
<string>l-mpyl 0.0 0.0 0.0 1 0 0 0 </string>
<string>r-mpyl 0.0 0.0 0.0 1 0 0 0 </string>
<string>l-mwing 0.0 0.0 0.0 1 0 0 0 </string>
<string>r-mwing 0.0 0.0 0.0 1 0 0 0 </string>
<string>*FLASHER* -0.5 0.0 16.8 0 2 0.0 0.2</string>
<string>*FLASHER* -0.5 0.0 16.8 0 1 0.0 0.1</string>
<string>*FLASHER* 0.5 0.0 16.8 0 2 0.0 0.2</string>
<string>*FLASHER* 0.5 0.0 16.8 0 1 0.0 0.1</string>
<string>*FLASHER* 0.0 0.0 -5.3 0 2 0.0 0.5</string>
<string>*FLASHER* 0.0 0.0 -5.3 0 1 0.0 0.25</string>
</array>
Might be worth a try. When I was flying a Merlin, there were no hives or fuel stations...however it was possible to turn it inside a docking bay and fly out again, as it's the size of a cargo pod
Screet
Posted: Thu Jun 18, 2009 6:45 pm
by Nemoricus
Okay, I stand corrected. I didn't know.
Let us know what happens.
Posted: Thu Jun 18, 2009 7:18 pm
by Screet
Nemoricus wrote:Okay, I stand corrected. I didn't know.
Let us know what happens.
The Merlin does not have problems with the fuel station ?!? However, it's extremely small. Even if the subentities would cause the ship to be 10 times as big as it really is, it should not have problems...
Need a larger ship, similar to the Caduceus, with subentities for another test, it seems.
I cannot imagine that this is a Caduceus-only problem...
Screet
Posted: Thu Jun 18, 2009 8:44 pm
by Thargoid
I recall a post in the fuel station thread about a mod'd a-wing or b-wing or something (one of the star-wars ships) that had fuel station problems due to the (home-brewed) addition of a turret...