"Um, Actually"

scifista42 said:

Um, actually: It's not possible to kill a full health Revenant with one super shotgun blast, because vanilla RNG will never allow all 20 pellets to deal maximum damage. Killing a full health Revenant with a single rocket is theoretically also impossible (160+128 damage < 300 hp), but I remember reading somewhere on this forum that it was once achieved anyway due to some weird engine bug, there was even a demo of it.

There's a TAS demo where in MAP07 of Doom 2, Arachnotrons are consistently killed with 2 rocket hits, through manipulation of the RNG. But that could just mean that the TASer merely picked the "right" time to fire, not that the RNG was actually hacked. Obviously not something easy/possible to do in real-time, though you can never be certain of the OCD level reached by someone who has played the same map 10000 times ;-)

Also: there's an upper limit to how many monsters and items can spawn on a map at a given time, assuming that RAM is maxed out and CPU time is not a concern (though that means a different limit for vanilla, for 32-bit and for 64-bit source ports, and also depends on the compiler, memory alignment etc.).

I was actually talking about this demo (RjY explained it below). :)

Arctangent said:

Um, actually, bosses are completely immune to splash damage, so barrels do nothing to them.

Um, actually, Barons of Hell (the shareware Doom boss) can be persuaded to infight by a well placed barrel.

Well, ahmn, Revenant rockets always home in their targets so I could make the rocket home me into a barrel and the Revenant would be sued by the baron.

Gustavo6046 said:

Well, ahmn, Revenant rockets always home in their targets so I could make the rocket home me into a barrel and the Revenant would be sued by the baron.

Um, actually, they don't always home in.

In addition, if a barrel blows up from the first hit, nobody will be "sued", due to how the blame chain works in vanilla Doom. It's more tricky to set up a barrel infighting situation than people think.

That old bug was hilarious, though, how the monster would hit a barrel and then target itself. "Did me just HURT me!? I'll frikkin' KILL me!"

Cacodemon's fighting
http://www.youtube.com/watch?v=O4RONetTmbs

scifista42 said:

If you take everything into account, then a hitscan attack has the fastest movement (infinite).

What about the Arch-vile "Fire" and "Zap" attacks and all the melee attacks?

Azuruish said:

Cacodemon's fighting
http://www.youtube.com/watch?v=O4RONetTmbs

Heh it reminds me, I once had made a test E1M1 meant to showcase the various barrel blowing modes. The player could move in and out of the view of various cacodemons, booby-trapped with barrels, blow the barrels himself, or tease the cacodemons to do it.

In one room, there was a single barrel exactly in the middle of two cacodemons, each hidden from each other's view (diagonally), but both within the blast radius of the barrel. Results:

• If the player blew the barrel himself with several pistol shots, of course the cacos would get angry at him.
  
• If the player moved so that one caco blew the barrel, then if it blew it in one shot, the other caco didn't retaliate.
  
• If it took more than one hit to destroy the barrel, then the one who damaged it last but without blowing it up, got the "blame" for it. In Doom v1.1, this could lead to one cacodemon "infighting with itself", even if it was the other one which destroyed the barrel.
  
• It was also possible for the player to wait until the barrel got hit once by one of the cacos, and then attempt to blow it up with ONE hit: if successful, the last caco to his the barrel would get the "blame" for the boom, not the player ;-)
  

In the other room, I simply had a chain of barrels and two cacos. No matter what or how, they never infighted due to a chain reaction. This can be explained by the barrel death code: it doesn't set its "killer" as its attacker, so "blame" doesn't propagate down barrel chains (some source ports have changed this behavior though). Don't ask me to track down the WAD or the original thread though....the Search function has been nerfed, WTF?!

Maes said:

Don't ask me to track down the WAD or the original thread though....the Search function has been nerfed, WTF?!

This seems to be the only other place you've mentioned it

Maes said:

Don't ask me to track down the WAD or the original thread though....the Search function has been nerfed, WTF?!

Um, Actually, https://www.doomworld.com/vb/newsearch.php?query=poster%3Amaes+barrel+chain+cacos

So this is really the NEW and ENHANCED Search function used by everyone? I thought it was some kind of a joke or that somehow I travelled back in 1999 (something about it reminds me the very early DW days). I was using the old, detailed one just days ago :-/

Well, FWIW the link to BARREL.WAD is dead. I gotta fish it from the Dungeon and put it back online somewhere else.

Edit: heh, it's been 4 years, what did I expect?

(According to Masters of Doom) During a 90s game convention, Doom II had a booth that was a paper-mache monster head that had dildos for teeth, and (Based on my own research) no pictures exist of this booth.

How much paracetamol did you take before you wrote that?

According to this article http://doomwiki.org/wiki/Armor_percentage_rollover, in first versions of Doom you can't get the negative armor percentage in practice.

EDIT: fixed link.

I'd get rid of that comma on your link, it sends you to a blank page. Just move it outside the url tags.

SavageCorona said:

How much paracetamol did you take before you wrote that?

Everyone wants to know. Trust me. Even if they day that they don't.

Cacodemons easily infight each other with no need of bites. And it is impossible to make them infight with bites.

SavageCorona said:

Um, actually, you didn't specify which group the Archvile belongs to, implying that Doomguy is included, which invalidates your statement, since Doomguy is the fastest moving thing in the game, BUT if you were to include non-living things such as PROJECTILES, then the plasma gun projectile is the fastest thing in the entire game.

Um, actually, Doomguy is faster than plasma with the help of wallrun, thingrun, turbo or damage boost.

Gustavo6046 said:

Cacodemons easily infight each other with no need of bites. And it is impossible to make them infight with bites.

Huh, what?

Maes said:

Huh, what?

You must go deeper to understand his post otherwise meaning will be hidden.
Amm... Interesting what a real speed of doomguy? Not in game units, in km/h or milles. Feels like doomguy runs within 20 km/h or 13 milles.

Or the maps feel bigger than they really are so Doomguy can reach one side of a small room in small time like in real life, but they are rendered as real big (read BUG) ones.

Azuruish said:

You must go deeper to understand his post otherwise meaning will be hidden.
Amm... Interesting what a real speed of doomguy? Not in game units, in km/h or milles. Feels like doomguy runs within 20 km/h or 13 milles.

Depends on how you map game units to real-life distances, but there's no single accepted way to do that. There are however unambiguous, absolute values for the speed in map units per tic [mu/tic].

If we accept that 16 mu = 1 foot, then there's the following equivalence ( a speed of 1 mu/tic becomes 1/16 feet per tic, or 35/16 feet per second [fps]. Since 1 mile is 5280 feet, 1 mu/tic = 1.516 mph). This gives us:

| Action     | mu/tics | mph   |
+------------+---------+-------+
|Walking     |  8.333  | 12.63 |
|Strafe-walk |  7.999  | 12.13 |
|Running     | 16.667  | 25.27 |
|Strafe-run  | 13.333  | 20.22 |
|SR-40       |  21.23  | 32.19 |
|SR-50       |  23.57  | 35.74 | *
|Wallrunning |  30.00  | 45.49 |
|SR-Wallrun| |  42.42  | 64.32 | *
+------------+---------+-------+

Of course, if you belong to the 10 mu or 8 mu = 1 ft. "school of thought", the speed in mph will have to be revised accordingly, but the mu/tic speeds are set in stone.

BTW, those speeds were measured by modifying a source port and measuring actual movement vectors (how many map units, using Euclidean distance, Doomguy moves in one tic). Walking/running and strafing were measured independently, except for SR-40. Wallrunning was only measured on perfectly N-S or W-E walls, not diagonally.

I wasn't able to reproduce SR-50 in order to measure it, so I computed it as the euclidean vector sum between the Running speed and (Strafe-Running * 1.25). Similarly, I assumed that a "strafe-wallrun" is simply the vector sum of two normal wallruns (at 45 degrees).

Either way, Doomguy can really run fast. Even with a 16 mu/ft assumption, his "walking" pace is already close to the average speed of a Marathon athlete (close to 13 mph), while his running speed (25 mph) is close to the average speed of a grand-tour cycling athlete, a 100m-dash champion or a moped. If you use 8 or 10 mu/ft. equivalence, those speeds could even double (leading to the infamous 50 mph speed claims).

BTW, Doomguy cannot sustain a speed greater than 30 mu/tic in any direction of the xy-axis regardless of HOW this speed is imparted (this includes explosions etc.), because of a built-in MAXMOVE speed-limiting constant in the engine (that is, regardless of floor friction). No matter how hardly you push him, he will only accelerate up to that vector-combined 42.42 mu/tic speed for one tic. He has, so to speak, a "terminal velocity" on the horizontal plane, no matter how hard he's pushed, even by stuff like extremely overpowered DEHACKED BFG fireballs and the such.

The only exception is movement in the z-axis: in that axis, he can acquire even greater speeds, and being Doom's gravity unrealistically strong, that leads to things dropping quite fast, especially when they are moving slowly.

Maes said:

Huh, what?

I think he meant provoking them to infight. You can't use their bites to provoke each other and they can be provoked without their bites, IE barrels.

MetroidJunkie said:

I think he meant provoking them to infight. You can't use their bites to provoke each other and they can be provoked without their bites, IE barrels.

Well, for that matter, the only monsters that could ever be provoked to infight using nothing more than melee attacks, were Demons and Spectres in Doom v1.1, because their attack worked more like a short-range projectile, and they could be tricked to misdirect it, thanks to the ridiculous delay. After removing this feature, but leaving the delay, we were left with pinkies uselessly running around, biting the air. In v1.1, such a "bite-in-the-air" attack would actually hurt anything that had moved in the (short) range. With post v1.1 melee, that isn't true anymore: nothing other than the intended target can be hurt, assuming that it didn't move away in the meantime.

Maes said:

Well, for that matter, the only monsters that could ever be provoked to infight using nothing more than melee attacks, were Demons and Spectres in Doom v1.1, because their attack worked more like a short-range projectile,

Um, actually it worked like a short-range hitscan.

Maes said:

| Action     | mu/tics | mph   |
+------------+---------+-------+
|Walking     |  8.333  | 12.63 |
|Strafe-walk |  7.999  | 12.13 |
|Running     | 16.667  | 25.27 |
|Strafe-run  | 13.333  | 20.22 |
|SR-40       |  21.23  | 32.19 |
|SR-50       |  23.57  | 35.74 | *
|Wallrunning |  30.00  | 45.49 |
|SR-Wallrun| |  42.42  | 64.32 | *
+------------+---------+-------+

Walking is MF25 command, so the max speed is 25/3 = 8.333. Strafewalk is MF24 so 24/3 = 8. Running max speed is MF50/3 = 16.667 etc. :-P Btw, wallrunning isn't 30.00. It is your current speed times 1.5 or 2 depending on situation. SR-50 maximum wallrunspeed is 23.57*2 = 47.14 (according to your table).

The fastest speed I have seen is: 30mu/tic north, 30mu/tic east AND you wallrun north and east which makes it 60mu/tic north and 60mu/east which equals to 60*sqrt(2) = 84.85. I think this is what happens in a void glide but it doesn't require that much of speed.

Looper said:

Walking is MF25 command, so the max speed is 25/3 = 8.333. Strafewalk is MF24 so 24/3 = 8. Running max speed is MF50/3 = 16.667 etc. :-P Btw, wallrunning isn't 30.00. It is your current speed times 1.5 or 2 depending on situation. SR-50 maximum wallrunspeed is 23.57*2 = 47.14 (according to your table).

As I said, all speeds except SR-50 and a "SR-Wallrun" are effective measured speeds, based on what I could actually perform. E.g. to wallrun, I only tried it on MAP01's NS corridor, which gave a repeatable and solid 30 mu/tic speed. To test max speed under external factors, I simply punched a barrel in E1M1. If I couldn't perform a test, I simply didn't report it.

BTW, I didn't realize the formula that linked effective speeds to commands was so simple: my understanding is that final effective speeds were the result of a complex interaction between friction, acceleration etc. (the player doesn't reach those speeds instantly, BTW). O_o

Now, for some reason I didn't leave the speed metering code in Mocha Doom, however to the best of my recollection, I put the metering somewhere in the ticker code, right before and after the player thinker was guaranteed to have moved, and printed the measured speed in mu/tic using the HUD. I should have kept that as a speedometer cheat ;-)

STFU PPL

Um, Actually speeds aren't interesting. We just have to kill a Cyberdemon. Now a BFG... RIPPIN AND TEARIN

Maes said:

BTW, I didn't realize the formula that linked effective speeds to commands was so simple: my understanding is that final effective speeds were the result of a complex interaction between friction, acceleration etc. (the player doesn't reach those speeds instantly, BTW). O_o

Yes, it is quite complex but the end result just happens to be 1/3 :O

Maybe I misunderstood what you meant with 'Wallrunning'. I just think that this is what happened: You used turbo and actually failed the wallrun, you got 30.000mu/tic and thought it was a wallrun. Getting 30.00 wallrun is actually really difficult (even under built if you add couple more digits). The minimum speed you need for a wallrun is 15.00000 so at that point the speed just jumps from 15 -> 30 but the speed is not capped at 30, it is capped at 60. Just like your SR-wallrun (strange name for diagonal running?) is capped at 30 but because it is diagonal, you get 30*sqrt(2).

I feel a bit confused and probably just didn't understand you and/or your wording D: