re: missile unit walk bug - Testers Wanted for 1.8.3 Fix

Post a reply


This question is a means of preventing automated form submissions by spambots.
Smilies
:D :) ;) :( :o :shock: :? 8-) :lol: :x :P :oops: :cry: :evil: :twisted: :roll: :!: :?: :idea: :arrow: :| :mrgreen: :geek: :ugeek:

BBCode is ON
[img] is ON
[flash] is OFF
[url] is ON
Smilies are ON

Topic review
   

If you wish to attach one or more files enter the details below.

Expand view Topic review: re: missile unit walk bug - Testers Wanted for 1.8.3 Fix

Re: re: missile unit walk bug - Testers Wanted for 1.8.3 Fix

Post by melekor2 » 22 Jul 2018, 20:31

Updated OP with link to the latest build thread.

Re: re: missile unit walk bug - Testers Wanted for 1.8.3 Fix

Post by melekor2 » 27 Jun 2018, 08:10

Thanks Spoon, added you on our forum now.

Re: re: missile unit walk bug - Testers Wanted for 1.8.3 Fix

Post by spoon » 26 Jun 2018, 15:27

Hey Melk

If you still need folks, my Tain forum name is spoon.

Cheers

Re: re: missile unit walk bug - Testers Wanted for 1.8.3 Fix

Post by melekor2 » 21 Jun 2018, 03:31

Myth has always been written partly in object oriented style (by Bungie), same as most C codebases.

If you're worried about keyloggers you better scan everything from 1.5.1 and up since I've worked on all of them.

Re: re: missile unit walk bug - Testers Wanted for 1.8.3 Fix

Post by portugay » 20 Jun 2018, 19:44

I read through that thread and I'm fucking amazed that you're adding object oriented c to this codebase.

Do I need to scan this shit for key loggers?

Re: re: missile unit walk bug - Testers Wanted for 1.8.3 Fix

Post by melekor2 » 17 Jun 2018, 18:57

If you had a problem with the link yesterday (Saturday) it should be fixed now.

Re: re: missile unit walk bug - Testers Wanted for 1.8.3 Fix

Post by melekor2 » 17 Jun 2018, 03:06

I should add: This fix does not affect warlock fireballs or fetch lightning, it only applies to projectiles that fly in an arc like dwarf bottles and arrows.

Since the issue appears most often when units are right at the edge of their range, the gameplay scenarios that are most affected are bowman on bowman and dwarf on dwarf battles. That would be my suggested focus for testing.

re: missile unit walk bug - Testers Wanted for 1.8.3 Fix

Post by melekor2 » 17 Jun 2018, 02:40

Back in 2013 I noticed this thing where missile units sometimes move forward a bit before attacking, even when they are already in range of the target. I made a post about it to get some feedback and most people seemed to be on board with changing this behavior. At the time I was planning to add it into 1.8.1 but after 1.8 was released I kinda forgot about it. Now I've finally got around to creating a fix, which I hope to add into 1.8.3. But I'm looking for people to test it out before we make that decision.

videos which show the problem:
https://vimeo.com/63034399
https://vimeo.com/63034400

the old thread:
http://www.myrgard.com/viewtopic.php?f=16&t=410

1.8.3 private beta thread & Beta Download here:
http://tain.totalcodex.net/forum/viewto ... f=9&t=7184

I don't anticipate any issues with this fix as it seems pretty straightforward, but in the end you never know. Myth gameplay can be subtle. Everyone is welcome to test, and I would specifically love to get some experts at competitive myth to give our beta build a rigorous test drive to make sure that this change is good for the game.

If you don't have access to our private testers forum, please post your tain username in this thread and I will upgrade your account permissions. PM might work too but I'm not sure if the email notification is working on here.



If you are interested in the technical details of the issue, I will try to give a quick description here:

The problem: When a missile unit fires, the projectile velocity is not always the same because there is a random component. The maximum range of the projectile depends on the velocity (faster = more range). When the game checks if the unit is in range or not, it uses the actual projectile velocity *including* the random component. So, if the unit rolls its random dice and comes up with a weak powered shot, its range for this shot may be less than for the previous shot, which causes it to start walking forward to get into range. The unit effectively has a kind of ESP/clairvoyance that its next shot is going to be weak.

The solution I've implemented: When evaluating the range, do not use the actual velocity from the random generator. Instead, use the average (expected) velocity. This way weak shots may just miss instead of causing the unit to walk forward. Note that such weak shots will not always miss, they may just end lower instead, hitting the target in the feet instead of the midsection.

Top