Page 15 of 54

Re: Split: Re-scaling experiment

Posted: Mon Oct 13, 2014 8:39 am
by Zireael
what? where? how do I test?

Re: Split: Re-scaling experiment

Posted: Mon Oct 13, 2014 9:09 am
by Redspear
Dizzy, thanks for the report and trunk updates.

Re torus drive:
Change the *2 to *4 for torus decelleration, that should do it.

Re nav beacon:
Thought about making it smaller (v. easy to do) and then I saw another adder next to it... it looked like a fly on a beach ball :-)

Re docking:
The adder shot was just outside the station, the angle may have been misleading. Sorry about that.

Zireael,
Not had a chance to tackle github yet (and not likely for next few days either), so the only way at present is to install from source, edit the code and then "make -f makefile" as signposted earlier in this thread. Then install the oxp.

Re: Split: Re-scaling experiment

Posted: Mon Oct 13, 2014 9:13 am
by Diziet Sma
Zireael wrote:
what? where? how do I test?
You need to obtain the Oolite v1.81 source code from Github ('Master' branch), edit 10 or so files in it as detailed here, then compile the code, install the OXP mentioned here, and then run the new executable..

Re: Split: Re-scaling experiment

Posted: Mon Oct 13, 2014 9:18 am
by Redspear
Re docking....

I think it's possible to set the docking bay (by which I mean the detection area, not the model itself) slightly further back in the shipdata.plist.
Need to be careful here though as the likes of the Anaconda are still full size...

Re: Split: Re-scaling experiment

Posted: Mon Oct 13, 2014 9:18 am
by Diziet Sma
Redspear wrote:
Re torus drive:
Change the *2 to *4 for torus decelleration, that should do it.
Thanks.. will give that a shot.

Re: Split: Re-scaling experiment

Posted: Mon Oct 13, 2014 9:23 am
by another_commander
I have uploaded a (totally unofficial) 64-bit build of Redspear's work for anyone who want to test but can't / don't feel like compiling source. Get it here. You will still need to download and install the OXP part and have the latest trunk installed so you can copy the exe over to its oolite.app folder..

Hopefully I've managed to follow all instructions correctly.

Re: Split: Re-scaling experiment

Posted: Mon Oct 13, 2014 9:24 am
by Redspear
A few lines below that is a decelleration factor that's applied only when mass-locked (*3 I think, in any case, the two are cumulative), so you could adjust that instead (if you like your asteroid fields to be hazardous ;) )

a_c, thanks so much :-D

Re: Split: Re-scaling experiment

Posted: Mon Oct 13, 2014 9:37 am
by Diziet Sma
another_commander wrote:
Hopefully I've managed to follow all instructions correctly.
If you have MIN_HYPERSPEED_FACTOR and HYPERSPEED_FACTOR set to 256, you've probably gotten it all correct.. :)
Redspear wrote:
A few lines below that is a decelleration factor that's applied only when mass-locked (*3 I think, in any case, the two are cumulative), so you could adjust that instead (if you like your asteroid fields to be hazardous ;) )
Ok.. I'll try that, cranked up to 4, first.. :D

Re: Split: Re-scaling experiment

Posted: Mon Oct 13, 2014 9:41 am
by Diziet Sma
Redspear wrote:
Re nav beacon:
Thought about making it smaller (v. easy to do) and then I saw another adder next to it... it looked like a fly on a beach ball :-)
That's actually why I thought of reducing the size.. my ballpark guesstimate of an appropriate size for a nav beacon would be about 1/2 the size of a Cobby3..

Re: Split: Re-scaling experiment

Posted: Mon Oct 13, 2014 9:51 am
by another_commander
For the first time, the Anaconda feels truly massive. Bring on the 750 tonnes!
(and ignore the scanner lollipops being off - I forgot to change the scale when I got the shot, but it's OK in the unofficial build).
Image

@Dizzy: Looks we're good, I used 256 for the two factors.

Re: Split: Re-scaling experiment

Posted: Mon Oct 13, 2014 10:58 am
by Redspear
Dizzy: ok, I can halve the beacon size and update the oxp (probably tomorrow).

a_c, did you use the torus decelleration that I originally posted or the one above?
If it was the former (my mistake, I forgot to re-make the source after my last tweaks and so it didn't show in last test), then take care flying out there folks :)

Re: Split: Re-scaling experiment

Posted: Mon Oct 13, 2014 11:38 am
by another_commander
Redspear wrote:
a_c, did you use the torus decelleration that I originally posted or the one above?
If it was the former (my mistake, I forgot to re-make the source after my last tweaks and so it didn't show in last test), then take care flying out there folks :)
I'm afraid I used the *2 deceleration, not the *4 one. Oh well, I'll try to upload another exe sometime later today or tomorrow then.

Re: Split: Re-scaling experiment

Posted: Mon Oct 13, 2014 11:44 am
by Diziet Sma
Redspear wrote:
If it was the former (my mistake, I forgot to re-make the source after my last tweaks and so it didn't show in last test), then take care flying out there folks :)
Aye.. whilst spectacularly entertaining for the locals (at least, for the ones who don't live too close to the action), punching a hole in a planet at 70 km/s is not good for one's long-term career prospects.. :lol:

Re: Split: Re-scaling experiment

Posted: Mon Oct 13, 2014 12:17 pm
by Redspear
Ah, hang on... the x2 is cumulative :|

So that means x2 for each part = x8 for testing the whole thing.

Dizzy, if you typed in * 2 to the code then you'll likely need * 8.
That would certainly explain your 'landing' :mrgreen:

If it still didn't work then try a total of x16.

Re: Split: Re-scaling experiment

Posted: Mon Oct 13, 2014 1:10 pm
by Diziet Sma
Well.. by following your suggestion, and leaving line 2670 of PlayerEntity.m at:

Code: Select all

float deceleration = (speed_delta * delta_t * HYPERSPEED_FACTOR * 2); // original has no * 2
and changing line 2676 to:

Code: Select all

deceleration *= 4; // original value 3
which would be a cumulative total of *8, I came to a (relative) halt at 0.1 planetary radius above the surface.. still rather scary, but safely above the danger point of 0.04-0.02.. sunskimming seems a little improved, too.. as in, I slowed down to normal speed shortly before the scoops activated.. as opposed to shortly after the tanks were full.. :lol:

I also stopped considerably quicker when mass-locked by ships.. no longer flying right by them before slowing down. Which is some pretty damn brutal deceleration, when you think about it.. :twisted:

If that holds true for subsequent tests, I'll be happy with that setting. Some folks might like a bit bigger margin, of course.


A couple of other observations..

Is it just my imagination, or is the Station Aegis smaller than normal? It seems like I get closer than before, before reaching the Aegis.

After running from a pack of pirates, I noticed I had to get them considerably beyond my scanner range before my alert condition changed from red to green again.. are NPCs able to maintain a lock beyond scanner range, or is their scanner range still the same size as before? IIRC, condition red is actually triggered by a NPC getting a missile lock, and I seem to recall you saying that missiles now have a range substantially larger than the scanner range.. related?