TFMurphy
Level: Smitemaster
Rank Points: 3118
Registered: 06-11-2007
IP: Logged

Didn't really want to make four different posts for four different bugs. I'm placing them in one thread so they can be ticked off and dealt with a little easier, without having to hunt them down in the various threads they've been mentioned. Anyways, here they are.

1. Entities die from falling off platforms at two main portions of each turn: after Mimics move and after all monsters have finished moving. However, due to the way the game currently checks for Mimics to have finished moving, any Mimic that has been placed after Clones or Decoys will move *after* the first check for entities falling off platforms. This is inconsistent.

2. Various things require updating to not go off on half-turns. We've discussed unhasted swords hitting Adders/Rattlesnakes at great length, so this is pretty much just one of the reminders. However, one thing only mentioned once or twice is that Aumtlich gazes break orbs every turn, which causes Speed Potions to break them twice as 'fast'. On top of that, Aumtlich gazes get a free 'break/crack orb' whenever a Player Double is placed [or when switching to another clone, answering a question, or when a builder removes oremites from underneath the player -- mrimer].

3. Brain Pathmaps need to be updated after Player Role changes and any time a Player Role loses stealth. (I've mentioned part of this before in a seperate thread, but this is the entire issue combined and restated for consistency)

4. drod.err's newlines aren't coming out correctly. "0D 0D 0A" is being written to the file instead of "0D 0A", which causes some text editors to hiccup over them. I've read that the likely cause of this is due to writing drod.err out in text mode, which would convert newline escape sequences already. Unsure what the best solution for this would be while still preserving the NEWLINE token in other non-file uses as well as on platforms other than Windows.

---

[Last edited by mrimer at 09-08-2007 04:40 PM]

Rabscuttle
Level: Smitemaster

Rank Points: 2539
Registered: 09-10-2004
IP: Logged

TFMurphy wrote:
Didn't really want to make four different posts for four different bugs. I'm placing them in one thread so they can be ticked off and dealt with a little easier, without having to hunt them down in the various threads they've been mentioned. Anyways, here they are.

Can't speak for anyone else, but I think it would be easier if they are in different threads - ticking off is done by the thread icon so can't be done individually when they are all in one thread. Also, if someone is searching for mention of a bug, a descriptive subject is more useful than a generic one. (That also bothers me in those 'list of 50 feature requests' that people sometimes come up with.)

Someone Else
Level: Smitemaster

Rank Points: 2398
Registered: 06-14-2005
IP: Logged

You can speak for me if you want.

silver
Level: Smitemaster
Rank Points: 915
Registered: 01-18-2005
IP: Logged

in as much as these forum posts serve double duty as bug tracking/progress topics, collecting unrelated bugs into one post is actually breaking the bug tracking system.


____________________________

---

[Last edited by silver at 09-06-2007 01:15 PM]

TFMurphy
Level: Smitemaster
Rank Points: 3118
Registered: 06-11-2007
IP: Logged

Well, I can delete this topic and post the bugs seperately, if people want.

When I posted, I was under the impression that the icon next to it was used mostly to remind the coders what issues were still outstanding, and once all four had been cleared up, it could finally be ticked and considered dealt with. At the time, I was thinking of multi-part issues all under a single related subject where some of the problems got fixed immediately, while others had to wait, so the bug topic was still considered open.

Rabscuttle's point about searching for bug topics is a very good one though, so I'll keep that in mind for next time.

Anyways, is there a deeper system beyond this forum that this kind of topic adversely affects? And whether there is or not, what solution would you recommend for this precise topic, given it currently exists? As said, I'll avoid collecting bugs like this again, so this topic should be the only problem now.

mrimer
Level: Legendary Smitemaster

Rank Points: 5462
Registered: 02-04-2003
IP: Logged

TFMurphy wrote:
Well, I can delete this topic and post the bugs seperately, if people want.

I don't mind having all these reports in one thread. Thank you for posting them.

---

[Last edited by mrimer at 09-06-2007 04:25 PM]

mrimer
Level: Legendary Smitemaster

Rank Points: 5462
Registered: 02-04-2003
IP: Logged

TFMurphy wrote:
1. Entities die from falling off platforms at two main portions of each turn: after Mimics move and after all monsters have finished moving. However, due to the way the game currently checks for Mimics to have finished moving, any Mimic that has been placed after Clones or Decoys will move *after* the first check for entities falling off platforms. This is inconsistent.

I've changed this logic so the first falling check occurs after all mimics, decoys, and clones have completed their turns. This shouldn't change anything else, since decoys+clones don't move anyway.

2. Various things require updating to not go off on half-turns. We've discussed unhasted swords hitting Adders/Rattlesnakes at great length, so this is pretty much just one of the reminders.

I've fixed entity sword hits thusly:

if (!this->bHalfTurn || bIsBeethroDouble(pMonster->wType))
   processSwordHit

However, one thing only mentioned once or twice is that Aumtlich gazes break orbs every turn, which causes Speed Potions to break them twice as 'fast'. On top of that, Aumtlich gazes get a free 'break/crack orb' whenever a Player Double is placed [or when switching to another clone, answering a question, or when a builder removes oremites from underneath the player -- mrimer].

I've augmented the aumtlich gaze API to take into account whether it's a half turn or not (which includes all those periphery cases listed above when no monsters are moving), and only break orbs on full turns.

3. Brain Pathmaps need to be updated after Player Role changes and any time a Player Role loses stealth. (I've mentioned part of this before in a seperate thread, but this is the entire issue combined and restated for consistency)

I'm now calling SetPathMapTarget when the player changes role and is targetable, and also when stealth is lost.

4. drod.err's newlines aren't coming out correctly. "0D 0D 0A" is being written to the file instead of "0D 0A", which causes some text editors to hiccup over them. I've read that the likely cause of this is due to writing drod.err out in text mode, which would convert newline escape sequences already. Unsure what the best solution for this would be while still preserving the NEWLINE token in other non-file uses as well as on platforms other than Windows.

Good suggestion. I changed writing the text in binary file mode and that fixed the problem!

TFMurphy
Level: Smitemaster
Rank Points: 3118
Registered: 06-11-2007
IP: Logged

Looks good.

mrimer wrote:
I've fixed entity sword hits thusly:

if (!this->bHalfTurn || bIsBeethroDouble(pMonster->wType))
   processSwordHit

<nod> That's the main thing (though I still disagree that inactive clones and decoys should get hasted turns, but that's a debatable issue that could've gone either way).

The problem I had when trying to fix it myself though was with consistency with other methods of swords appearing on half-turns: namely sworded NPCs appearing/spawning on a half turn due to scripting. It's a niche case, and a slightly tricky one to work through, but it seemed more natural to me to allow an extra sword hit in those cases than have them appear and not cause damage. That said, it wouldn't be the only case where a sword appears and not causes damage, but it'd likely be more common (and possibly slightly more useful, since with continuous Appears and Disappears, you could simulate a hasted NPC) than the exceptions I can think of.

Still, the general case was the main thing, and that's dealt with.