These forums have been archived and are now read-only.

The new forums are live and can be found at https://forums.eveonline.com/

Test Server Feedback

 
  • Topic is locked indefinitely.
12Next page
 

August Release - Missile Code Rewrite

First post
Author
Previous Topic Next Topic
CCP Lebowski
C C P
C C P Alliance
#1 - 2016-07-08 10:54:57 UTC  |  Edited by: CCP Lebowski
Good afternoon testing denizens of New Eden!

As of yesterday's update to Singularity, almost all missiles and bombs in the game (fired from capsuleer ships, citadels, POS batteries or NPCs) have been moved over to new shiny and hopefully more robust code! Practically this should have no effect on the functionality of missiles, aside from fixing the odd bug or two. However, we ask you all to be extra vigilant and report here if you find any problems. As always, I'll keep this thread updated with any issues found.

Known Issues:

  • None


Thanks and happy testing!

CCP Lebowski | EVE Quality Assurance | Team Five-0

@CCP_Lebowski

CCP Lebowski
C C P
C C P Alliance
#2 - 2016-07-08 10:55:05 UTC
Reserved

CCP Lebowski | EVE Quality Assurance | Team Five-0

@CCP_Lebowski

Amak Boma
Dragon Factory
xX SERENITY Xx
#3 - 2016-07-08 18:29:26 UTC
why rorqual got no love yet
elitatwo
Zansha Expansion
#4 - 2016-07-09 14:58:05 UTC
I have one. I was flying a Drake and the heavy missiles have an explosion radius of 246m (fury) and it clearly should be 187m and the explosion velocity is 104m/s and should definitely be 147m/s.

For the test I didn't use a computer or enhancer, just trying to see flaws on them first.

T1 missiles are having a similar problem. A nova heavy missile - tech one, no faction - should have an explosion radius of 102m and an explosion velocity of 212m/s.

I didn't use crash even though I had signs of withdraw. Can you fix that please?

Eve Minions is recruiting.

This is the law of ship progression!

Aura sound-clips: Aura forever
CCP Lebowski
C C P
C C P Alliance
#5 - 2016-07-13 17:59:18 UTC
Thanks for the report. When comparing the stats of missiles between current Tranquility and Singularity I'm not seeing any of the differences you are explaining above.

Could you link the fit you are using to run these tests?

Thanks!

CCP Lebowski | EVE Quality Assurance | Team Five-0

@CCP_Lebowski

Cade Windstalker
#6 - 2016-07-13 19:56:42 UTC
Out of curiosity, and because this was recently explained to me, are we still going to see the weird behavior in the last second of flight time where missiles sometimes expire before hitting a target at the listed max range and sometimes after it, depending on which server tick they expire on?
Cristl
#7 - 2016-07-14 01:32:59 UTC
Cade Windstalker wrote:
Out of curiosity, and because this was recently explained to me, are we still going to see the weird behavior in the last second of flight time where missiles sometimes expire before hitting a target at the listed max range and sometimes after it, depending on which server tick they expire on?

I think what you're asking is actually by design. Do you mean how missiles with, say, a 3.7 flight time travel for 3 seconds 30% of the time, and for 4 seconds 70% of the time?
Kalmr Varys
State War Academy
Caldari State
#8 - 2016-07-14 04:05:51 UTC
Cristl wrote:
Cade Windstalker wrote:
Out of curiosity, and because this was recently explained to me, are we still going to see the weird behavior in the last second of flight time where missiles sometimes expire before hitting a target at the listed max range and sometimes after it, depending on which server tick they expire on?

I think what you're asking is actually by design. Do you mean how missiles with, say, a 3.7 flight time travel for 3 seconds 30% of the time, and for 4 seconds 70% of the time?


I really hope that isn't by design. I think it's more a limitation of the tick system, since right now the tick system is 'all or nothing' and has no concept of doing partial amounts even if the item should have expired part way thru the tick.
Vesan Terakol
Trollgrin Sadface
Dark Taboo
#9 - 2016-07-14 05:16:25 UTC
Quick question on a related manner - does the new code adjust the situation, where you fire off a volley and warp out and the missiles disappear mid-flight?
Cristl
#10 - 2016-07-14 08:32:32 UTC
Kalmr Varys wrote:
Cristl wrote:
Cade Windstalker wrote:
Out of curiosity, and because this was recently explained to me, are we still going to see the weird behavior in the last second of flight time where missiles sometimes expire before hitting a target at the listed max range and sometimes after it, depending on which server tick they expire on?

I think what you're asking is actually by design. Do you mean how missiles with, say, a 3.7 flight time travel for 3 seconds 30% of the time, and for 4 seconds 70% of the time?


I really hope that isn't by design. I think it's more a limitation of the tick system, since right now the tick system is 'all or nothing' and has no concept of doing partial amounts even if the item should have expired part way thru the tick.

Erm...what? Yes it's by design, and it's designed that way because of the tick system Big smile

How else would they do it? What's your (hamster-friendly) suggestion?
CCP Lebowski
C C P
C C P Alliance
#11 - 2016-07-14 16:45:41 UTC
Thanks for the responses. All known issues with this rewrite should now be resolved, so let me answer some of your questions.

Cade Windstalker wrote:
Out of curiosity, and because this was recently explained to me, are we still going to see the weird behavior in the last second of flight time where missiles sometimes expire before hitting a target at the listed max range and sometimes after it, depending on which server tick they expire on?
Yes, this behaviour will remain the same with this new code. This is unfortunately just the nature of making missiles work in a system where a 1 second tick is the smallest amount of time possible!

Vesan Terakol wrote:
Quick question on a related manner - does the new code adjust the situation, where you fire off a volley and warp out and the missiles disappear mid-flight?
This is intended. Missiles guidance systems deactivate when unable to communicate with the ship that fired them (When it leaves their local area of space) Big smile

CCP Lebowski | EVE Quality Assurance | Team Five-0

@CCP_Lebowski

Valterra Craven
#12 - 2016-07-14 17:13:36 UTC
CCP Lebowski wrote:
Thanks for the responses. All known issues with this rewrite should now be resolved, so let me answer some of your questions.

Cade Windstalker wrote:
Out of curiosity, and because this was recently explained to me, are we still going to see the weird behavior in the last second of flight time where missiles sometimes expire before hitting a target at the listed max range and sometimes after it, depending on which server tick they expire on?
Yes, this behaviour will remain the same with this new code. This is unfortunately just the nature of making missiles work in a system where a 1 second tick is the smallest amount of time possible!


What if you added some "prediction" to the code? ie what is the likely probability that the missile should or should not hit the target in the last second?

Valterra Craven
#13 - 2016-07-14 17:15:00 UTC
Another question: What is the practical affect of this new code?

I.e. Is it 10% less intensive on the server cpu, etc?
Cade Windstalker
#14 - 2016-07-14 18:44:07 UTC
Valterra Craven wrote:
CCP Lebowski wrote:
Thanks for the responses. All known issues with this rewrite should now be resolved, so let me answer some of your questions.

Cade Windstalker wrote:
Out of curiosity, and because this was recently explained to me, are we still going to see the weird behavior in the last second of flight time where missiles sometimes expire before hitting a target at the listed max range and sometimes after it, depending on which server tick they expire on?
Yes, this behaviour will remain the same with this new code. This is unfortunately just the nature of making missiles work in a system where a 1 second tick is the smallest amount of time possible!


What if you added some "prediction" to the code? ie what is the likely probability that the missile should or should not hit the target in the last second?



Pretty sure they could do this, but it would involve basically rewinding the state of the missiles and the target, which would turn missiles into the lag monster from planet "bugger the servers" since at best you'd basically be simulating a partial tick for both the missile and its target inside a regular server tick.

I was hopeful that it would be fixed, but not terribly expectant.
Ix Method
Doomheim
#15 - 2016-07-15 07:56:08 UTC  |  Edited by: Ix Method
CCP Lebowski wrote:
Thanks for the responses. All known issues with this rewrite should now be resolved, so let me answer some of your questions.

Cade Windstalker wrote:
Out of curiosity, and because this was recently explained to me, are we still going to see the weird behavior in the last second of flight time where missiles sometimes expire before hitting a target at the listed max range and sometimes after it, depending on which server tick they expire on?
Yes, this behaviour will remain the same with this new code. This is unfortunately just the nature of making missiles work in a system where a 1 second tick is the smallest amount of time possible!

In that case shouldn't the numbers just be rounded off and missiles balanced accordingly? Losing DPS cause you trained something to IV instead of leaving it at 0 is a ludicrous situation tbh.

Travelling at the speed of love.
Cade Windstalker
#16 - 2016-07-15 14:27:12 UTC
Ix Method wrote:
CCP Lebowski wrote:
Thanks for the responses. All known issues with this rewrite should now be resolved, so let me answer some of your questions.

Cade Windstalker wrote:
Out of curiosity, and because this was recently explained to me, are we still going to see the weird behavior in the last second of flight time where missiles sometimes expire before hitting a target at the listed max range and sometimes after it, depending on which server tick they expire on?
Yes, this behaviour will remain the same with this new code. This is unfortunately just the nature of making missiles work in a system where a 1 second tick is the smallest amount of time possible!

In that case shouldn't the numbers just be rounded off and missiles balanced accordingly? Losing DPS cause you trained something to IV instead of leaving it at 0 is a ludicrous situation tbh.


Do you have an example where this would be the case, because I can't think of one...

This only happens if the last tick of travel is less than one second, so you should never lose DPS because you trained something higher, because without that training you wouldn't be able to hit the target at all.

CCP could potentially either make missiles always round up (instead of rolling to see if the missile is alive at that last tick or not) but that would, in may circumstances, make hitting those break points over powered and make anything past that redundant. They could also make missile flight times always be a flat number of seconds and remove anything that affects that by a percentage and either change it to a flat time increase (in a round number of seconds) or move those bonuses to velocity increases.

Personally I'm a bit on the fence here, since while this behavior is unintuitive and should probably go away just boosting missile velocities instead of that and flight time seems less than idea, as does the flat rounding...
Valterra Craven
#17 - 2016-07-15 14:57:24 UTC
Cade Windstalker wrote:


CCP could potentially either make missiles always round up (instead of rolling to see if the missile is alive at that last tick or not) but that would, in may circumstances, make hitting those break points over powered


Exactly how does making a missile hit a target in the last second always, vs sometimes and sometimes not hitting it become over powered?
Cade Windstalker
#18 - 2016-07-15 15:45:04 UTC
Valterra Craven wrote:
Cade Windstalker wrote:


CCP could potentially either make missiles always round up (instead of rolling to see if the missile is alive at that last tick or not) but that would, in may circumstances, make hitting those break points over powered


Exactly how does making a missile hit a target in the last second always, vs sometimes and sometimes not hitting it become over powered?


Sorry, I think I may have been unclear.

I meant that if that becomes the case hitting break-points with skills or modules becomes over-powered, not that missiles themselves do. For example if I get 5% into the next "tick" of flight time anything past that is worthless, because I don't actually gain more flight time until I get over another full second.

Past that point rigs, skills, modules, or anything else that increases flight time is pointless, because the only bonus that really counts is the one that gets you over the last second, if everything rounds up.

That's why I proposed making missiles have either fixed flight times or make all bonuses to flight time full second bonuses, which removes the issue of bonuses not mattering until you hit the next breakpoint, but causes its own issues in other ways with how missiles would have to be balanced, especially short-range ones.
Valterra Craven
#19 - 2016-07-15 17:41:18 UTC
Cade Windstalker wrote:
Valterra Craven wrote:
Cade Windstalker wrote:


CCP could potentially either make missiles always round up (instead of rolling to see if the missile is alive at that last tick or not) but that would, in may circumstances, make hitting those break points over powered


Exactly how does making a missile hit a target in the last second always, vs sometimes and sometimes not hitting it become over powered?


Sorry, I think I may have been unclear.

I meant that if that becomes the case hitting break-points with skills or modules becomes over-powered, not that missiles themselves do. For example if I get 5% into the next "tick" of flight time anything past that is worthless, because I don't actually gain more flight time until I get over another full second.

Past that point rigs, skills, modules, or anything else that increases flight time is pointless, because the only bonus that really counts is the one that gets you over the last second, if everything rounds up.

That's why I proposed making missiles have either fixed flight times or make all bonuses to flight time full second bonuses, which removes the issue of bonuses not mattering until you hit the next breakpoint, but causes its own issues in other ways with how missiles would have to be balanced, especially short-range ones.


I think you are going a little too into your own logic. Breakpoints in and of themselves cannot be over powered. Hitting a breakpoint or not doesn't mean that everything that came before it or after it is meaningless. In the case of missiles the only important aspect is if the missile does damage or not. Doing damage at 55km and not doing damage at 56km or 55.1km or 55.000001km is not something that you abuse. A use case that is already so rare with little meaningful gain either way can not be overpowered.
Blitz Hacker
Bored Bastards
From Anoikis
#20 - 2016-07-15 19:45:22 UTC  |  Edited by: Blitz Hacker
Cade Windstalker wrote:
Out of curiosity, and because this was recently explained to me, are we still going to see the weird behavior in the last second of flight time where missiles sometimes expire before hitting a target at the listed max range and sometimes after it, depending on which server tick they expire on?
Yes, this behaviour will remain the same with this new code. This is unfortunately just the nature of making missiles work in a system where a 1 second tick is the smallest amount of time possible![/quote]
In that case shouldn't the numbers just be rounded off and missiles balanced accordingly? Losing DPS cause you trained something to IV instead of leaving it at 0 is a ludicrous situation tbh.[/quote]


Why don't they just make flight time whole seconds and adjust velocity of missiles instead vs flight duration. Then you would always get intended distance, and would never 'miss' a flight at/near max range. velocity speed would only be modified and the dictator of distance reduce the BS due to 'server ticks', also would go with the making missiles more viable in pvp (as presumably this would lead to increased missile velocity and less 'lag/delay between pressing fire and impact)
12Next page