Whackanoob Dialogue Suggestions
 

Some may remember an old minigame I had within the GC Studios called Whackanoob, which was basically a version of Whackamole where you'd hammer newbies, some of which would be saying irritating/frustrating/comical things.

This is now being remade into a proper studio event. If anyone has any suggestions for some funny things for the newbies to say feel free to post here or let me know.

I currently have:

Poor spelling/grammar is encouraged.

graal support

pls rate gp

ill hack ur account
my friend is gm i can get u bant
how do i join gp
noob!

U MAD?

n# y u no whack? (I think n# is the one that shows nickname and not account name.)

Trololo

lookn 4 gst team!!

A noob wouldn't say this and even if they did, they're right.

Good suggestions.

Not sure why I didn't think of this.

You'd be surprised how many people have been parroting remarks about the hit detection simply because they're aware of how terrible it was in the past, and in some cases aren't even aware the old serverside one was scrapped. It has also been adopted by certain newer players, sometimes through mimicking, sometimes by defaulting to excuses after successive losses, in the same way someone might accuse someone who beat them of hacking.

As for whether it's right or wrong, nobody can ever come up with a reasonable explanation as to why they believe it is wrong. The vast majority of the time it's "I hit them on my screen and they didn't get hurt" and sometimes "the people watching the spar saw me hit them too", which is ofcourse not how default hit detection is supposed to work, as it's a matter of your opponent seeing you hit them on their screen.
I've even heard ridiculous reasons ranging from "the steps.wav sound in the gani isn't the same as the default walk gani" to "the arena is the wrong size".

It would be good if someone can correct me if I'm wrong, but my understanding is that default HD scans the gani/sprite of other players on your client, and if they're slashing, checks an intersect of 32 x 32 pixels offset by their direction from their body's blocking area, with a 32 x 32 pixel box of your own body's blocking area. I have an online debug level displaying polygons over these boxes which matches perfectly with an offline equivalent level (hence default).

It is also my understanding that default recoil moves the player at 1 tile distance per frame according to the hit angle, for 5 frames. That blink lasts between 0.55 - 1.95 seconds (again, good if someone can correct me if I'm wrong) according to release of held arrow keys or slashing. Recoil is also sticking players to walls.

The only very slight difference I can admit to, is that our movement speed is pixel rounded, whereas default level 3 shield speed being at 0.6 tiles per frame is not pixel perfect.

Unless I were to actually see the client's code or somebody were to offer a credible explanation as to why they think it might be wrong, there's nothing more I can do to possibly make it more similar, other than some speculative experimentation with what order things might be occurring with default, or if player slashes are scanned for multiple times within the same frame.

I already told you what was wrong ingame. Hitting people from a mile away is not normal and nothing of the sort happens on UN or iClassic with or without TCP. I don't know if it's the movement system or hit detection, but from a player perspective it doesnt really matter; something is off.

Going by the Clientside mechanism Default works by, it's perfectly plausible to hit somebody from a far distance according to your own perspective. If the time it takes for data to go from your computer to your opponent's computer is high enough (such as against someone with a ping into the thousands), you could even hit somebody from the opposite side of the arena according to your own perspective, provided their client detects an intersect.

When you have 2 Europeans sparring eachother using TCP, there's always going to be a big delay between slashes going from one to the other, as the data is having to travel all the way from Europe to the USA, and then back to the opponent in Europe. A decent UDP transfer among 2 Europeans would change the situation entirely, as data would then not be having to travel so far to and from 2 Europeans.

I've seen videos of spars on UN and iPhone, tested default early in Classic's development as well as pre-wipe, opened up Dev Polo plus an arena (using default) on one occasion where most of the servers were down, and even sparred against a crowd of people on that place which is considered a taboo, guessing ahead of time to successfully hit an opponent inwhich there's a delay seems completely normal to me.

. . .

Classic's current hit detection is weird under certain circumstances where it isn't on UN, iClassic, or other "default" servers in the past. I don't really know what else to say, but something is off about it.

As a result of the default level 3 speed being 0.6 tiles per frame, a value which can not be divided into whole pixels, your characters display is rounded to the nearest whole pixel coordinate, which is also the value synchronised to the server and other players, while your actual clientside x/y values remain at decimal pixel values. Because of this, with default it is actually possible to see your character edge-on with another players sword, but you don't get hit because your absolute x/y value could be up to half a pixel away from that visible edge. The opposite can also be true, the synchronised value may be closer to your opponent than your absolute x/y value, which can result in your opponents client detecting a hit even though you were really up to half a pixel too far away.

That is the only difference I can account for, and if anything Classic's method is more accurate, as the visible coordinate matches the absolute coordinate on your client.

On the other hand, there have also been those with experience who have been complimentary of the new system, who've said it feels right. When developing it I would even swap it around with Default and people who were on PKing couldn't tell the difference, though that may well have been the fact it was level 1 shield speed of 0.5 tiles per frame, which is pixel perfect.

I would have to agree with Crono, I think thats why Classic was never really a huge sparring server. The hit detection is very awkward and weird especially if you're used to the other 'default' servers, it almost makes it frustrating.

Thor is right, what Crono described used to happen all the time on default. You would spar someone on 28.8k or 56k modem (or they were intentionally lagging) and the only defense for this was to swing where they were going which made it appear as if you were swinging across the screen.

Incase this is hard to understand, perhaps this picture from Dev Polo (default) will explain it more easily:

http://img155.imageshack.us/img155/5648/pixeldif.png


So what's weird about it compared to default exactly?
There's nothing more I can do unless someone can actually explain it, and it's not another case of "I hit people on my screen and they don't get hurt".

Go look at Java Graal's source and port the hit detection over.

Already looked at it, and I did see a small snippet of code Stefan had shown to Dan several years ago, but it only really confirmed what I had already suspected.

Except it doesn't happen on other servers using the "default" system (nowadays that's really only UN and iClassic, neither have this problem).

Look at this hit Blobz's opponent received on 17 seconds.

Horrible example @ Blobz. Sparring people I've sparred on UN and iClassic on Classic felt completely different. The HD just reacts differently. I mean, I can't believe how obvious this is...

I don't hit people from awkward distances on either server, only on Classic. Maybe it's the level 3 shield thing, I don't know, but as an active player who actually plays multiple servers I can confirm that it's not the same. The only time it felt "real" was when we both sparred eachother with UDP on.

Whoever neg-repped me for like 1 point can **** off, it's not a "placebo effect".

How is it a horrible example?
It clearly shows his opponent being hurt a whole 2-3 seconds after Blobz had slashed around that area. This backs up the idea that with the combination of default hit detection and lag a sparrer has to guess ahead of time.

It's not obvious though, a lot of sparrers do not notice a difference, and those that claim to can either never explain it, or have some pre-conceived notion based on pre-wipe Classic's HD.

This is either due to some technicality involving the different machines the servers are hosted on, sheer coincidence, or a biased exaggeration. This behaviour is perfectly normal (even if confusing) to occur as a result of the default clientside hit detection mechanism, and is exactly what the video shows. There is nothing scripting wise on Classic that would be invoking a higher delay, synchronised variables are kept to a minimum where they might be used more wrecklessly on an older more dated server.

The movement speed being pixel perfect does mean it is not exactly the same as default, however the difference is only very slight (visually un-noticable), and is also more accurate.

You know what, I can go all day about how it's not the same (from first-hand experience, not being biased, etc) but I'm not going to bother. It's not what this thread is intended for anyway.

:rolleyes:

I don't really know how Crono can explain it more than he already did. On Classic you can hit someone from farther away than other servers, you really can't boil that down to anything more simple than that.

This isn't like an attack against your scripting, but something as a community (majority) would like to see changed.

this thread sucks

That's a description of a symptom, not the underlying problem, and doesn't help much in trying to fix it. As Thor has said, it would help if someone could present some specifics on how to reproduce the problem (e.g. a specific case where Classic's HD registers a hit that the classic system's doesn't).

Where would you start in trying to fix this problem? There are a number of possible causes, and that's assuming that the problem even exists. It's like chasing ghosts to try to fix it when it can't be reliably reproduced. There's no way for him to know if a change he makes resolves the problem because it's impossible to test (it's based on the "gut feeling").

I always hate when I'm in the position Thor is in right now. There's really nothing he can do, but the players keep complaining. I suspect he could change back to classic HD without telling anybody and still get the same complaints.

Why don't you run the hit detection/movement simulation with two NPCs while simultaneously using the same input with the default system and see if they go out of sync at any point? That'd be really cool to check out and it would give you definitive results.

Don't you think if we knew exactly what the problem was we would tell him? I mean we obviously want to see it fixed here, and we really don't have the privileges to check out scripts and certain things to troubleshoot.. and that is why we were trying to provide 'symptoms'.

What position is Thor in? I think you're over-thinking a little too much Chris, I'm merely just trying to get a point across about a symptom.. not to bash a very helpful developer.

I've already debugged the 2 strenuously, for one example in the form of an offline and online NPC showing a 32 x 32 pixel polygon to represent boxes, and concluded that the basic hit mechanisms are like-for-like. I've even shown this to complainers and they've either become convinced, or rejected it with some belief that player damage checks are scripted differently to showcharacter() ones.

Certain people having this gut feeling that there's a problem but without being able to explain it isn't a good enough justification, such gut feelings have an unreliable tendency and often contradict the gut feelings of other people.

A short while ago, one of the most experienced sparrers on Graal said something like "this hit detection is better than UN's", which didn't make any sense, as to my best knowledge Classic's system is emulating default.
More recently, during the lag problems, they changed their mind to "this hit detection is worse than UN's", and were quite vehement in the suggestion that I must have changed the hit detection, when the system has been the same for years.

The problem with this alleged symptom is that it is perfectly normal behaviour for the default hit detection, and is therefore expected to occur within our scripted system. The suggestion that it does not occur eleswhere simply isn't true (see video), and even if it was, could easily be explained by circumstances such as the recent server lag problems, or being hosted on an inferior machine, rather than the scripting being off.

Unless someone were to provide a measurable suggestion, or I were to see all of the related client code, experimenting with changes would be like chasing ghosts as Chris said.

Next time there's a Classic spar tourney I will use Fraps and point out hits that shouldn't count. I'll upload the vid on YouTube. It's not something subtle either, it's incredibly obvious and other sparrers agree with me. I can't tell you why it's happening because I'm not a scripter, I'm just a player.

Lovely neg-rep @ the guy who repped w/ "I'm Crono and I ***** when ever someone proves me wrong.", even though I'm not even close to being wrong or proven wrong. lol

So do what I said, and then nobody will complain any more. Enable default movement and your movement at the same time, but instead of overwriting the player's XY coords just use a showgani and see if they ever stop overlapping one another. Have a full spar with this on to test out edge cases and the way lag behaves.

The concept behind it being a clientside hit detection means that when it comes to attacking other players, it is essentially irrelevant what you see on your own screen, and what does matter is where your opponent would see you slashing on their screen, where it takes varying time for your data to reach them.

Making such a video and pointing out instances where you've hit a player from long distance according to your own perspective would mean absolutely nothing.

If other sparrers agree with you that you shouldn't be able to hit players from long distance according to your own perspective, then there is a massive misconception at hand. It is an entirely plausible occurrence which can and does happen with default systems, as shown within the video I posted before.

I did something similar to this back in 2009 when I was making the scripted system. Now we have a special server option in place which completely disables default aspects which cannot be done through script.

But I'm very confident there is nothing wrong with the sword detection mechanism itself, I made this video to demonstrate:



The only thing I'm not certain about is what order default sword checks are occurring in relation to movement, such as before, after or both. I need Stefan to confirm this.

How would it mean nothing when it would show something that only occurs on your server and not the others? lmao seriously I give up, your system only benefits my sparring style anyway.

You posted a video with Blobz, an infamous lagger. He had well over 2-3 seconds of latency. Poor example.

Now you're contradicting yourself.

By continuing to post, yeah. Other than that I don't see where I contradict myself.

\/

Also the video was simply the first video which showed up in the search results.

An incredibly laggy player will have messed up HD, but that wasn't the case while I was sparring on Classic now was it? Of all players you had to post Blobz aswell, I think at one point his UN ping reported over 5000ms? I'm comparing the HD from my perspective between UN, iClassic, and Classic where only Classic has funky hits land while the other two don't for me.

You stated that these delayed hits do not happen on other servers in response to Maximus's post concerning laggers, therefore you shouldn't be surprised when I show an example of it happening as a result of a lagger.

As for the validity of your claim that it's a common occurrence on Classic, me being a European hosting the spar tournaments I'd have been seeing your opponents get hurt at a similar interval. I didn't see them being hit from miles away from you, with the exception of one person who actually did have a high ping and was running at you like a headless chicken, so no surprise there.

Extreme laggers have always been exceptions and unpredictable in terms of hit detection: be it the bare default system, Classic's old serverside HD, or Classic's current system. My claim that these hits don't happen was geared towards the normal player. I was landing hits I don't think I should have been landing (hits which wouldn't have landed on the other servers) all throughout the tournies vs various players and my ping was consistent (100/10 connection without any hickups). I still spar on iClassic and UN so it's not like I'm out of touch with hit detection or anything.

The only time there were no issues was when the two of us sparred UDP.

Dialogue:"showadims"

Keep in mind that the reason the spar tournaments were being hosted in the first place was because all the main servers had crashed. Classic may have not suffered a crash, but there were definitely some lag problems in patches whenever this happened.

This is good.