View previous topic :: View next topic |
Author |
Message |
Watermelon has entered the game!
Joined: 08 Nov 2012
|
Posted: Fri Apr 12, 2013 1:28 am Post subject: Blockmap fix dmflag? |
|
|
After some conversation on #zdplayers, ZD uses as much vanilla as it can. Therefore I assume the blockmap bug was put back in to be vanilla. One debate that many of us have had whether it exists or not is something with the ssg acting weird (believed to be dropped shots). I think this may be due to the blockmap... which would make sense since it's definitely not the unlagged.
If the blockmap bug is implemented to act like vanilla, can I request a DMFlag for us to use that can turn it off? |
|
Back to top |
|
|
phenex2 Unstoppable!
Joined: 10 Jan 2008
|
Posted: Fri Apr 12, 2013 3:15 pm Post subject: Re: Blockmap fix dmflag? |
|
|
Watermelon wrote: | Therefore I assume the blockmap bug was put back in to be vanilla. |
We did not put anything 'back in', because ZDaemon at no point used ZDoom's modified blockmap code to begin with. We always used the original vanilla behaviour.
ZDoom changed the blockmap behaviour on October 29, 2003 after the 1.23b33 release. |
|
Back to top |
|
|
Watermelon has entered the game!
Joined: 08 Nov 2012
|
Posted: Fri Apr 12, 2013 5:53 pm Post subject: |
|
|
Therefore is it possible to request a quick patch w/ a DMFlag for it and test it out to see if it causes the issue we've been running into?
CTF wasn't designed to be strictly vanilla (looking around/RJ's, slopes, flags, ACS...etc etc). The blockmap can cause critical shots to go blank which can determine the outcome of a game. |
|
Back to top |
|
|
phenex2 Unstoppable!
Joined: 10 Jan 2008
|
Posted: Fri Apr 12, 2013 9:50 pm Post subject: |
|
|
Watermelon wrote: | Therefore is it possible to request a quick patch w/ a DMFlag for it and test it out to see if it causes the issue we've been running into? |
If the blockmap would be the cause for the 'issue' then you would have noticed it since the first ZDaemon 0.8 release. I thought you are claiming this is some 'new' issue in 1.09.
Before we even consider to add such a deep engine change for no good reason i would like to see some evidence that an issue actually exists. You claim that the SSG drops a high amount of shots because of the blockmap bug. If this were true then it implies that the the same must be true for the railgun which is much easier to verify. SSG demos are of no use because of its randomness. I would like to see a demo where railgun shots of the recording player that clearly hit the target are not registered as hits by the server. So far i have not seen a single demo that shows anything suspicious that would indicate the existence of an issue.
If its not possible to produce such a demo with railguns then the 'issue' you are seeing is most likely just the randomness of the SSG.
What should be kept in mind when recording such a demo currently is this bug though. Just try not to move the mouse while spawning or teleporting.
Watermelon wrote: | The blockmap can cause critical shots to go blank which can determine the outcome of a game. |
I think you overestimate the frequency and impact of the blockmap bug. The blockmap bug only occurs when target and shooter stand perfectly aligned very close to a blockmap boundary and are very close together. Given the size of a blockmap cell and the randomly distributed player positions in a 'real' game the chances to run into this bug are very small. Also both teams are equally likely to be affected by the bug so it has neither a positive nor negative impact on the overall outcome of the game in the long run. |
|
Back to top |
|
|
Madgunner Dominating!
Joined: 23 Nov 2003 Location: Atlanta,GA
|
|
Back to top |
|
|
phenex2 Unstoppable!
Joined: 10 Jan 2008
|
Posted: Sun Apr 14, 2013 11:57 am Post subject: |
|
|
Thanks madgunner. As you can see in that article a VERY specific setup is needed to trigger the bug. You don't just run into this bug in a normal game like that. I am sure that some experts like memfis or dew can give us some insight about how often this happens in SP for example.
The effect of the blockmap bug is indeed that a shot passes through the enemy, but what really happens to players in game is that they either did not aim well enough or they have the false expectation that the ssg does guaranteed onehit kills from a certain distance, which can simply NOT be assumed given its randomness. |
|
Back to top |
|
|
dewww God like!
Joined: 29 Dec 2006
|
Posted: Sun Apr 14, 2013 1:11 pm Post subject: |
|
|
here comes the blockmap bug again, i feel like i've posted this several times at several places already, heh.
this issue is overrated.
yes, you will trigger it pretty much every time you fight the mastermind with hitscan weapons (ssg, bfg tracers, ...), but i'm sure this is not the context we're talking about. mastermind is so big it will always stand over different blocks of the map and yes, then some hitscans may get lost if they hit the bastard over the wrong block.
but we're talking about small targets, right? well, another common single player occurence is shooting from melee distance, which is exactly what madgunner links to. because unlike fists, chainsaw is a hitscan weapon too. :) if you're figuratively hugging a monster (or even your opponent... *sniff*, respect despite rivalry!) you may whiff your whole ssg load (eww...) through their sprite, because they're standing over a block edge and you connect all your hitscans in the "wrong" block. but this is really only noticeable when literally touching the other thing! the effect becomes less and less noticeable with distance growing between you and your target very soon, because it's less and less probable that you will connect your pellets in the wrong block. here's a bullshit equation:
Code: | Ibb = (a / d^3) * (-1)^m |
where Ibb is importance of the blockmap bug (measured in giveashits, a less known SI unit), a is the area the target occupies in units squared, d is distance (yes, it's cubed) and m stands for mode and equals 1 for DM and 2 for SP.
in a context where locking yourself against your opponent in a touching distance is not a realistic scenario and it's improbable that all your ssg pellets would connect in the wrong block, the whiffing issue is probably connected to something else (ping, rng, skills, etc.). i can imagine bfg tracers getting lost that way, but no one ever brings that up it seems. you do know the zandronum bfg is wildly overpowered compared to vanilla, right?
EDIT: clarification of some silly sentence constructions. |
|
Back to top |
|
|
Kilgore Air Cavalry
Joined: 17 Jun 2003 Location: Up the river
|
Posted: Wed Apr 04, 2018 4:35 pm Post subject: |
|
|
The "blockmap bug" mentioned by watermelon was really a problem with the older version of unlagged. It was also called "dropped shots" and attributed to packet loss by others. This flaw in unlagged was discovered & verified right after the 1.10 release and has been fixed as of 1.10.01; therefore, there is no need to mess with the engine. |
|
Back to top |
|
|
|