Forums » Suggestions

Increase the priority for the collision damage retune.

Nov 23, 2013 draugath link
The title pretty much says it all.

Please increase the priority level for retuning collision damage.

It's been mentioned in the past that the current state of collision damage is a side-effect of fixing a long-standing bug in the physics engine. That's all well and good, but why hasn't collision damage, an arguably integral part of the game mechanics and experience, been retuned to match it's previous levels?
Nov 23, 2013 TheRedSpy link
Worst bug in VO. Hands down.
Nov 23, 2013 greenwall link
because the bug hasn't been fixed??
Nov 23, 2013 draugath link
TheRedSpy is calling it a bug to be dramatic, when that is not the case here. A bug was fixed that wound up decreasing the amount of damage applied on impact with the existing collision values, as previously stated by Incarnate. What needs to happen is that collision damage now needs to be retuned to coincide with the fixed physics engine.
Nov 24, 2013 Conflict Diamond link
TheRedSpy? Dramatic?

/me faints

/me lifts head long enough to say:

+1 to Draugath

/me faints again
Nov 24, 2013 incarnate link
I've actually bumped the priority of this substantially, we are going to look at it as soon as we can. But we're kinda buried under "stuff". This is not a trivial thing to "fix", we need to understand exactly what the ramifications were, and go back and analyze the differences in the code paths, and probably do some A/B testing to get a clearer idea of what happened. Collision-response to an iterative impact over time, is not a super trivial topic; a lot of different things can impact the simulation. And It doesn't seem like the sort of thing where we should just randomly change damage numbers willy-nilly and "hope it feels similar".

Among other things, crap keeps happening. Like on Sat (yesterday) morning our office firewall crapped itself, killing our VPN and any remote-development; so I spent most of the day today rebuilding that, and things are still not quite 100% yet. Fires happen, and they have to be put out, and they detract from other development.
Nov 24, 2013 Conflict Diamond link
You boys need to write a Blues song this month. Or a Country/Western one, but I'm not a fan so make it a Blues song.

If when you can look at this, can you consider a threshhold for "paint-scratch" collisions to cause no hull damage? so Tridents don't read 94% when they are really at 99.999% after someone spooged their landing, and pet bots don't attack stations you scrape while docking? (funniest thing ever, next to them attacking an asteroid I bumped, until one of them shot the other and they all 3 exploded)
Nov 24, 2013 abortretryfail link
Even if it doesn't feel similar. Vendetta Online: Nerf Edition is silly. I personally liked the wingbug mode better. :(

Inc, check this out if you're looking for a new firewall: http://www.pfsense.org/
It's easy to configure, backup, and restore. Runs on commodity PC hardware. Firewall died? grab another old junk PC, stack it full of PCI network cards, boot it up off the CD with your config file on a USB stick and you're back in business. :)
Nov 25, 2013 incarnate link
If when you can look at this, can you consider a threshhold for "paint-scratch" collisions to cause no hull damage? so Tridents don't read 94% when they are really at 99.999% after someone spooged their landing, and pet bots don't attack stations you scrape while docking? (funniest thing ever, next to them attacking an asteroid I bumped, until one of them shot the other and they all 3 exploded)

I agree with that, small impacts should basically be non-events. And preferably not even trigger the red-flash or anything.

Inc, check this out if you're looking for a new firewall: http://www.pfsense.org/

Thanks, I am aware of pfsense; not what we use, but we did roll with FreeBSD based firewalls for many years. I've been into "pf" for so long, I used to custom-compile Darren Reed's original ipfilter ("ipf", the precursor to pf) for Solaris 2.6 and FreeBSD 4.x. It is by far my favorite filtering syntax.

Unfortunately, fast-flash and simple config loads don't help when the 3-year-old backup of the config goes AWOL, and you have a lot of weird specialized routing and rulesets :(. Such is life.

If I were really serious about it, I'd roll a pair of freebsd machines and do redundant hot-failover via CARP; but we've gone out of our way to make our office not all that mission critical, so it just hasn't been worth the effort (although perhaps I should re-evaluate that, heh).
Nov 25, 2013 TheRedSpy link
The change to collision damage thingy is an example of a phenomenally bad call between two broken systems. Exploding in disproportionately minor wing clipping scenarios is nowhere near as game-play damaging as the ability to fly at-will inside asteroids.

I'm not sure its been demonstrated to you guys at guild just how easy it is to slip inside asteroids during combat.

I try not to make feature videos on youtube about bug exploitation but this one is asking for it. I once won a conquerable station by hiding in a big asteroid against like 8 other people and the only ships I havent managed to get inside an asteroid is the XC and the Type-M. Some of them, particularly the vultures, you can get in and out of any asteroid thats big enough to fit them in seconds. Sometimes you fly right though an asteroid during combat like it wasn't even there just by accident. The old way where you had to be careful of you'd blow up was far far better both strategically and in terms of no. of parts of the game that were broken.

Why can't we just go back to the old way in the interim. I remember this was explained, but I damn sure don't remember buying the reason.

Me? Dramatic? NEVER!
Nov 25, 2013 abortretryfail link
I'm not sure its been demonstrated to you guys at guild just how easy it is to slip inside asteroids during combat.

... if you happen to be laggy all the time. I've got a ~60ms ping most of the time and getting inside an asteroid is pretty difficult. On my phone's LTE it happens all the time.

If I were really serious about it, I'd roll a pair of freebsd machines and do redundant hot-failover via CARP;

Inc, the reason I suggested pfSense is because it's just about as powerful as a roll-your-own FreeBSD firewall box and takes a hell of a lot less time to set up and manage. I've been running it in production at work since 2007 and our IT dept has been understaffed since 2006. I'm sure you can relate. :)
Nov 25, 2013 incarnate link
The change to collision damage thingy is an example of a phenomenally bad call between two broken systems. Exploding in disproportionately minor wing clipping scenarios is nowhere near as game-play damaging as the ability to fly at-will inside asteroids.

But the "wingbug" is NOT why we changed things; you make a lot of assumptions, and then I spend a lot of time trying to back-track those wrong assumptions. Anyway, I think at this point your opinion has been duly noted, and I've already said we'll look at it as soon as we can.

Why can't we just go back to the old way in the interim. I remember this was explained, but I damn sure don't remember buying the reason.

The new system is threaded, and spreads the physics/collision load over many cores. This is especially important on the server. Similarly, the client and server collision behavior must be identical or it'll create a lot more problems. I don't really care if you "buy" that reason or not.

I'm not sure its been demonstrated to you guys at guild just how easy it is to slip inside asteroids during combat.

... if you happen to be laggy all the time. I've got a ~60ms ping most of the time and getting inside an asteroid is pretty difficult. On my phone's LTE it happens all the time.


Yeah, it's always been possible to get inside asteroids with a bad ping, it's just less dangerous to try it now.

Inc, the reason I suggested pfSense is because it's just about as powerful as a roll-your-own FreeBSD firewall box and takes a hell of a lot less time to set up and manage.

Yup, like I said, I'm familiar with it, from the very first origins of the project. But the distance between "just about as powerful" and "as powerful" becomes more significant as your configuration gets weirder. What we use now is "just about as powerful" too, and it works for the moment, I just couldn't lay hands on a config backup.