Page 2 of 2

Re: EotA: Twilight 1.14g6 Download

Posted: September 21st, 2013, 12:11 am
by DarnYak
You ruined my night. Thanks.

Also I see you lurking there. Drop by wc3 or steam if you've got a few could use a little elaboration.

DarnYak

Re: EotA: Twilight 1.14g6 Download

Posted: September 21st, 2013, 7:17 am
by Executor
My experience with Maw of Death is it either always works or it will rarely work in a game. What I mean is that consuming multiple units will only result in 1 spit when it doesn't work.

Edit: I've been getting fatal errors today (4 times). Prince Rue was in each of those games, but he might have nothing to do with it.

Re: EotA: Twilight 1.14g6 Download

Posted: September 21st, 2013, 1:49 pm
by Scheba
DarnYak wrote:From an older thread:
•Maw only gets one charge for digesting/spitting regardless of number of units devoured (I think I observed getting two charges once but couldn't reproduce it)
This seemed to work fine when I tested it, but it could use a bit of clarifying. Assume you eat 3 units. This gives you 3 corpses to spit, but there's a delay between spits (1.667 seconds) - the mana number indicates the number currently available for instantly casting, but here its one at a time. So if you spit only one, and wait 2 seconds it should jump back up to 1, for a total of 3 casts. It also leaves the most recent order in a queue for when it refills, so if you click twice it'll cast immediately upon the second becoming available. If it still appears broke with this clarification, please let me know.
The problem was not only the number of projectiles, even if I pressed Esc to digest the units, I only received the healing amount for one charge. I noticed this problem in an online game of version 1.14g5 and reported it after testing it in a local game (to make sure I hadn't just observed it wrongly and that it wasn't network lag related either) and being able to reproduce it. When I repeated the test for Brainlust in version 1.14g6 (and found it to be still broken), I also tested Maw again, but it worked correctly, so I thought the bug introduced in version 1.14g5 (or maybe a little earlier) was fixed.

Now, the recent post by Executor suggests a different scenario: that there are some things that sometimes work correctly for the entire duration of a game and sometimes don't work correctly even once during a game. The problem with Chili Breath that watermelon reported seems to fall into this category as well (I wasn't able to reproduce it when I tested it a while ago). Some reports of strange AI behaviour may also be cases.

It might be that these problems have a common cause. Unfortunately, as long as this cause (or these causes, however many they may be) remains unknown and unfixed, this means that observations may not be reliably reproduceable, which makes testing more time-consuming.