zex20913
Level: Smitemaster

Rank Points: 1721
Registered: 02-04-2003
IP: Logged

File: demo bug.JPG (149.9 KB)
Downloaded 82 times.
License: Public Domain

There's now a bug in rooms with clones, without conquer tokens. Observe screenshot. My "score" on CNet is 41. My actual score (not counting clone switches) is 37. It's optimal (I think) so I'd like to keep that #1 score.

____________________________

---

[Last edited by zex20913 at 04-09-2007 04:32 AM]

NiroZ
Level: Smitemaster
Rank Points: 1302
Registered: 02-12-2006
IP: Logged

Screenshot? Where?

zex20913
Level: Smitemaster

Rank Points: 1721
Registered: 02-04-2003
IP: Logged

It's attached. I had it in PNG at first, and altered it to JPG.

____________________________

nobodybutyours
Level: Delver
Rank Points: 50
Registered: 06-22-2005
IP: Logged

This is the only room with clones that I've completed so far, so I can't generalize yet, but unless there's something wrong with this room only, I assume the problem has to do with whether or not switching control counts as a move. Perhaps the in-game engine doesn't count it as a move, but the spider checking the demo does?

Some observations:

--Switching control to/from a clone, does count as a move as far as the undo queue is concerned. (i.e., move beethro and switch control to a clone; if you then hit the undo key it switches control back but will not undo the prior move).

--I see the same thing as zex looking at my demos now. In fact, it surprised me to see that my solution was initially alone at #1, considering how simple the room is, so it's no surprise that it's now considered the same score as all the prior #1s.

--Maybe Zex does have a more efficient solution, since mine was 38 moves (showing up later on the uploaded list as 40). Zex's screenshot shows 37.

--The first time I went through the room, I did it in 39 moves, which the demo info reflected, but it was later changed in the uploaded list to, I think 41.

--So something odd is going on.

--The exact discrepancy in my cases (adding two), strongly suggests it has to do with the spider counting the two Tab-switches between Beethro and his clone as moves.

---

[Last edited by nobodybutyours at 04-09-2007 05:48 AM]

nobodybutyours
Level: Delver
Rank Points: 50
Registered: 06-22-2005
IP: Logged

UPDATE: I went and completed the room south of the one in question ("twice south").

The demo first appeared both in the local and online lists as 75 moves, still shows up in the local list as 75 moves, but now, a few minutes later, shows up in the online list as 79 moves.

Indeed, including the initial clone potion activation, I had switched between Beethro/Clone exactly four times (79-4=75). (another bit of possible evidence: when watching a demo, the switch between clones and Beethro counts as a move as far as using the arrow keys to advance/rewind the demo is concerned).

UPDATE2: Zex, I see your demo showing up on the list for the first room ("once south") as 41 moves (tied for 2nd). Any chance you hit tab three or four extra times just to see how the switch plays out?

Everything seems consistent now except for Zex's initial 37 move score.

(EDIT: Ah, I just figure out how to do it in 37. It saves a move by switching back to Beethro, moving once square toward the exit (making the roach move into clone-striking position), switching back to the clone, striking, switching back to Beethro, and exiting. So, indeed, counting the clone-control switches, it comes out to 41)


UPDATE3: The issue is actually generalizable beyond clones. In fact, any time a player takes one of a number of actions that counts as a "step" as far as demo playback is concerned but does not count as a "move" as far as the game is concerned, the action is erroneously counted as a move in the online tally. Such actions include placement of mimics and placement of decoys.

---

[Last edited by nobodybutyours at 04-09-2007 06:44 AM : update3]

WyattSmee
Level: Goblin
Rank Points: 19
Registered: 11-14-2006
IP: Logged

nobodybutyours wrote:
Such actions include placement of mimics and placement of decoys.

Correct. Mimic/decoy placement bumps up the counter by one, as Beethro did physically move into the square containing the potion. An interesting thing about clone placement is that the spider currently counts this as 2 moves. One for Beethro and another for the forced initial switch to the clone's perspective.

As for the mismatched values in zex20913's demo (Duration 37/CaravelNet 41), fortunately this discrepancy seems to be fixed in the upcoming patch, based on the green checkmark next to the similarly-affected Conquer Token bug thread.

I'm still unclear myself on whether Caravel intend for clone switches to be counted towards room solution scores. Personally I'd vote yes for the sake of switch optimality. And somebody really ought to think about standing up for monster rights. Things start to get a bit unfair with Beethro zipping around like this.

---

[Last edited by WyattSmee at 04-09-2007 07:36 AM]

coppro
Level: Smitemaster
Rank Points: 1308
Registered: 11-24-2005
IP: Logged

No, because that means that clicking a clone is better than tabbing to it.

KevG
Level: Smiter

Rank Points: 333
Registered: 08-16-2004
IP: Logged

I just checked, clicking on a clone ups the move count the same as tabbing does. I personally don't care whether or not switching counts towards the high score. I'd just like to know one way or the other what the final verdict is going to be. For example, my current score in the conquer token/clone room is 68 score/52 moves, but I know how to do it in 67/53. I'm assuming that the first one is going to be best once the patch comes, but it would be nice to know for sure.

WyattSmee
Level: Goblin
Rank Points: 19
Registered: 11-14-2006
IP: Logged

coppro wrote:
No, because that means that clicking a clone is better than tabbing to it.

Aha. That would be so true. An excellent point, and one I hadn't considered.

Briareos
Level: Smitemaster

Rank Points: 3516
Registered: 08-07-2005
IP: Logged

KevG wrote:
I personally don't care whether or not switching counts towards the high score.

I'd say no, as with 3 or more clones in a room clicking would be a definitive advantage over tabbing then, which is rather silly...

np: cLOUDDEAD - Rifle Eyes (Ten)

____________________________
"I'm not anti-anything, I'm anti-everything, it fits better." - Sole
R.I.P. Robert Feldhoff (1962-2009)

zex20913
Level: Smitemaster

Rank Points: 1721
Registered: 02-04-2003
IP: Logged

Amen to that, guys. Movecount should be the basis, and since there are two ways of switching between clones which are unequal in "moves", it shouldn't matter. Fortunately, this is what was intended.

____________________________

nobodybutyours
Level: Delver
Rank Points: 50
Registered: 06-22-2005
IP: Logged

Correct. Mimic/decoy placement bumps up the counter by one, as Beethro did physically move into the square containing the potion. An interesting thing about clone placement is that the spider currently counts this as 2 moves. One for Beethro and another for the forced initial switch to the clone's perspective.

Looking over my demos, there's the same discrepancy between local result and online reported results for rooms with mimics and decoys. I'm pretty certain the spider is counting placement of those as two moves as well. . . (edit: these were all from days ago, so perhaps that had already been addressed and fixed)

(In any case, shouldn't the fix be a CaravelNet one, independent of client-side patches/updates?)

(edit: With clicking instead of tabbing, it looks like my 37 for 1S is standing as 37 on the online list (in-game). Obviously that should go to Zex, or whoever found that solution first)

---

[Last edited by nobodybutyours at 04-09-2007 04:44 PM]

JoMax
Level: Delver

Rank Points: 65
Registered: 12-28-2005
IP: Logged

Completed TCB AC 3S1E on Apr 16 with a move count of 192, went back to it and achieved 175 moves (1st place!). I noticed 2 days later that captainzakku (G-Spot) had beaten my high score by 1 move (Well done!) so I went to download his demo.

In the CaravelNet window on the Demos screen my recorded demo was shown as 186 moves, checked the stats on the web pages which verified the 186 moves (expected). Thinking there may have been a problem with spider (6400+ unverified demos) I re-completed the room with 175 moves, which was initially recorded on CaravelNet as such and now after verification is shown as 186 moves, and captainzakku's as 187!

This is the only time I have noticed a difference between my demo and that recorded by CaravelNet

[EDIT] This is the same problem as Spider and clones? Move count difference is due to clone/mimic/decoy interaction!

---

[Last edited by JoMax at 04-20-2007 03:48 PM : Duplicate problem]

krammer
Level: Smitemaster

Rank Points: 904
Registered: 02-12-2003
IP: Logged

This is probably due to the bug reported here.

____________________________
+++ Divide by Cucumber Error +++ Please Reinstall Universe and Reboot +++

"Oook!"

jemann
Level: Master Delver

Rank Points: 160
Registered: 01-13-2006
IP: Logged

Oh good - this bug has been reported. I couldn't figure out why demos for the AC level (with all the clones) were being 'inflated' by the spider.

It seems to me that the in-game system is the right one, and that clone-swapping shouldn't count as a move - if only because the monsters fail to react in the meantime. IMHO, of course.

mrimer
Level: Legendary Smitemaster

Rank Points: 5056
Registered: 02-04-2003
IP: Logged

I just noticed I hadn't checked off this bug thread as fixed, but I think I reported that this will be fixed in 3.0.2 somewhere...

____________________________
Gandalf? Yes... That's what they used to call me.
Gandalf the Grey. That was my name.
I am Gandalf the White.
And I come back to you now at the turn of the tide.