tain.totalcodex.net

Orlando the Axe wrote:I'm pretty sure that Tai Kung ended up being right about 1.6 trow pathfinding being altered. Correct me if I am wrong Iron.

This is not true, click here. At the bottom of the original post it will say the following:

(As a historical note, the Trow Polka is what prompted my flame earlier in the MWC. V 1.3 also behaved that way. I was completely off base about it!)

So after long and many posts about it, this is the only thing mentioned about him admitting being wrong.

As for who helped test, I don't recall Tai ever testing other than the first 1.7 beta. Qwerty, well something tells me he wasn't that useful if he did test. The other 3 I wouldn't know if they had either way.

Yes Tai was wrong about the trow polka being a post-1.3 bug. Nonetheless, 1.7 fixes this bug.

His other two main complaints were (IIRC) herons turning around to heal themselves (another 1.3 bug) and what I termed the "bloodlust" bug, where if you've got a bunch of melee units fighting & tell them to move away, some units will refuse & keep swinging.

The bloodlust bug was caused by fixing a similar (but inverse) 1.3 bug, where if you told a bunch of moving units to attack, some would refuse & keep moving - ie. the annoying-as-hell, cost-you-a-vital-MWC-game "archer walk bug". The bloodlust bug is fixed in 1.7.

Oh, and I believe the heron-turning bug's fixed as well, though Myrd would have to verify that as I believed my preliminary fix for it caused problems elsewhere.

We've also revisited the magic dirt problem, though I don't know if the fix is as good as the 1.5.0 version. Basically if a unit with a large radius (eg a trow) was close to other units, it would get stuck & unable to move, even if facing directly away from the other units. The fix was in looking to check whether the unit was indeed facing away, and if so not checking to see if there was a collision. I really don't know why it stopped working between 1.5.0 and 1.5.1, and without 1.5.0's sourcecode its impossible to find out.

OtA hits the nail on the head with his summary

About the pathfinding being altered......

I was told that upon investigating the "trow pathfind bug" in the code for pathfinding they noticed an incorrect value entered of -1 instead of the correct +1. what exactly was involved i dont know. this is not responsible for the precived trow whatever bug, but is something they found while checking to make sure everything WAS working properly with the pathfinding code.

It was a Magma member that told me this well over a year ago, Doobie i belive.

I keep mentioning this in diffrernt places on the forums and so far no magma people have responded. Would be nice guys! I blame this incorrect value for the NEW way that units ALLWAYS try to intercept the enemy unit where the enemy IS GOING TO BE and NOT just run AT THEM like they allways did in 1.3

When this started i dont know, but i swear i never saw it in 1.5 or anything before that.

If people need films of this i can make em, its really easy to spot it happen when you know what to look for.

people are able to exploit this issue with with great success in multiplayer, so it would be nice to be sure it is indeed fixed and i imagine that would of been done when they found the error. :::crosses fingers:::

-Renwood

Ren, that's a different bug from the "trow polka" one, and it affects all units.

Imagine if you have a warlock inside a formation of, say, 10 myrks in a loose formation, all in guard mode so they can't move. An enemy trow walks up to kick the warlock, and it _looks_ like he should have enough room to penetrate the formation. Instead he gets reasonably close then starts walking back & forth without trying to walk in. That's trow polka, and its fixed in 1.7 by adopting the, would you believe, TFL pathfinding algorithm for large units.

What you're talking about is units leading their target, walking towards where the enemy is going to try and head them off instead of directly following.

The mistake you make is that this is a 1.3 feature. Myth 2 has always had units lead their targets in this way. TFL did not, it had units walking to where their targets were without any "intelligence", which led to quite different melee strategies especially vs giants.

Bungie made some errors in their leading code which led units to occasionally get it very wrong in deciding where their targets were headed. We spent a lot of time with Acheron in particular investigating this in both 1.3 and the patch at the time (1.5.1?). I believe a fix is in 1.7 that improves Bungie's code - Myrd could verify that. We are not reverting it to TFL's behaviour, unless you've got vTFL on of course. I don't know about your -1 instead of +1 thing, doesn't ring any bells for me.

Iron, i gotta say i dont dout what you say is true, but i used to play this game like a religion years ago.

I used to be as l337 as the rest of them and a very good judge of unit behaviour. I swear this was never in bungie made versions of myth 2.

Ive been noticing it for about the last 3 years or so, and it allways really bugged me because i have to BABY SIT my units when i tell them to attack somebody. Say a enemy warrior is chasing and hacking your dorf parrallel from east to west.

Now your styg is only like 2-3 world units away and when you expect your styg to run AT the enemy war and start chasing behind him and hacking him (like i belive all bungie myths allways did) your stys will INSTEAD run parrallel from east to west just like the warrior and dorf are doing.

So you HAD plenty of time to save the dorf but the stupid styg WOULD BEARLY go twards the war, and is instead happy to just run where the warrior is going to be! im SURE the dorf will be allready dead by then.

So now what you keep having to do to prevent the death of the dorf is NOT click on the enemy warrior chasing the dorf, you have to instead select the styg and keep clicking on the ground in the direction OF the enemy war. because if you click on the war your styg will start just doing the intercept at a point BS. Now when your in the middle of a battle controlling a buncha units, and you KEEP having to babysit units beacuse the pathfinding is fuxord and they cant do a simple thing like chase or RUN AT ENEMY AND SAVE DORF, this makes something that should of taken ONE CLICK OF THE MOUSE into a clickfest leaving you unable to spend time clicking ALL YOUR OTHER UNITS! beacuse 1 damn unit cant do what he was allways able to do in bungie myth's TFL AND SB.

I know this game pretty well, sure im rusty at playing it i dont spend 10 hours a day online killing people in FFA and giant team games like i used to.

But i know for a fact that this isnt the original myth 2 or 1.3 bungie unit behaviour.

You saying exactly what Acheron said 2/3 years ago, and he was just as insistent that we'd changed it until we did some testing with him in 1.3 & saw exactly the same behaviour.

Sorry, real testing beats memories/perceptions every time, no matter how strongly you express them. I'd be much more interested to hear whether you've looked for this bug in 1.7 & whether its better or worse than 1.6.

So THERE Renwood! (I dunno)

I am not sure exactly what changed or when - but magma did something that changed how myth II plays in one or more of the updates. The reason I knos is because back when I played 1.3 for 5-6 hours a day, I was a WHOLE LOT better at Myth. Seriously, I can't put my finger on what's different now but over the past few years I only play a few hours a week, but things feel different than I remember and I really hope Magma is going to fix 1.7 so I can be good at myth again.

Thank you.

ps. I HAV FEELMZ.

I was never good at Myth. Does that count as a bug? Can you fix it?

No i havent been able to check for this in 1.7 as i need somebody to playtest it with me to recreate it.

great you guys are going to force me to install diffrent versions and get somebody to play them with me so i can set up the scenes to show this happen

Id think my 11 years of playing/creating this game would garner more faith in me from people. maybe just a little? hehe

If i end up being inncorrect in the end ill buy a peach and not eat it.... as punishment, For my crimes of missremembering shit.

But wait, how can i just imagine it was like this with pathfinding when Doobie confirmed this when i asked him about it long ago on hl.

No need to install anything. Just put the Myth II 1.3 exe in the Myth folder and take out all the Patch plugins that come after that version. And since it will be 1.3, I suggest you take out as many plugins you don't plan on trying, because it will load them all before going to the main menu. Oh, and take out that Myth 2 CD or make an ISO file and mount it since it will need it to run the game.

Yea thats the problem my myth 2 cd is 10 years old and has a tiny hole in it. ~8^(

Renwood TWA wrote:Id think my 11 years of playing/creating this game would garner more faith in me from people. maybe just a little? hehe

Unfortunately memories are fallible. Old memories even more so. Old memories in old brains even more so. So you are down like three strikes in your 'gimme a little faith' campaign.

A-Red wrote:I was never good at Myth. Does that count as a bug? Can you fix it?

Not for you sadly. However all my units in multiplayer games will get 200% health with the 1.7 patch.

im pretty sure that trows used to be able to take down 10 berserks before dying, not they barely kill 4-5. you guys completely screwed over 1.3. I was celest during most of 2003, i know what i'm talking about. i was pro and you guys all are non-playin newbs. I heard from a guy I dont know if he does any developping for myth, maybe he was just messing with me to shut me up, but he told me they are dividing some variables by 0 in the pathfinding code for trows.