Please tell me I'm not the only one to ever do this

**will_asher** · April 4, 2011, 06:51

Done that a few times (once or twice with greater vaults, at least a couple other times with lesser vaults).

On another note, it seems to me that a greater vault without a single artifact shouldn't happen (unless you've found them all).

**Timo Pietilä** · April 4, 2011, 07:16

Originally posted by will_asher

On another note, it seems to me that a greater vault without a single artifact shouldn't happen (unless you've found them all).

There is a bug in 3.2.0 that doesn't guarantee excellent objects in "8" blocks of GV:s. It is one word fix in generate.c

This is what it should be:

Code:

				/* Nasty monster and treasure */
				case '8':
				{
					place_monster(y, x, p_ptr->depth + 40, TRUE, TRUE);
					place_object(y, x, p_ptr->depth + 20, TRUE, TRUE);
					break;
				}

This is what it is:

Code:

				/* Nasty monster and treasure */
				case '8':
				{
					place_monster(y, x, p_ptr->depth + 40, TRUE, TRUE);
					place_object(y, x, p_ptr->depth + 20, TRUE, FALSE);
					break;
				}

When I got a CGV in 3.2.0 and best weapon it had was some quarterstaff of slay orc which might have actually been drop from OoD monster it was too much for me and I fixed that one.

**Nick** · April 4, 2011, 07:49

I was desperate enough once to let an Intermittent Software Bug infestation to get out of hand in O - they blink, teleport, tunnel and breed. Eventually they let me into the vault containing the dungeon book I wanted, and I got breathed to death by a Balor within two grids of the book.

**ulrichvonbek** · April 4, 2011, 13:22

There were plenty of ego items...Maybe I just got unlucky in that none of the excellent passed the artifact test.

**fizzix** · April 4, 2011, 16:45

Originally posted by Timo Pietilä

There is a bug in 3.2.0 that doesn't guarantee excellent objects in "8" blocks of GV:s. It is one word fix in generate.c

I'm not sure this is a bug as much as it was a design choice, by someone, for reasons I dont know.

**Timo Pietilä** · April 4, 2011, 17:00

Originally posted by fizzix

I'm not sure this is a bug as much as it was a design choice, by someone, for reasons I dont know.

Nobody seem to know. Maybe greater vault reward was considered too high at some point, but this change just plain rendered them complete waste of time. Kind of like someone considering that weapons of westernesse are too good and decides to give them -10 speed to compensate.

**Scraper** · April 5, 2011, 05:11

This raises a question I have about vaults. Is it possible to blink/ teleport etc in to them?

After battling the witch king, who teleported me away, I found myself just below the vault but with no easy access. I would have had to have gone a long way out of my way to get back to the entrance. I didn't want to do that as I had TOed many of the nastier nasties and didn't want to have to face them again.

I decided to try to blink into the vault. No success after many tries. In the end, I dug my way around the vault and close to the original corridor leading to the vault . Cast blink and bam, straight into the corridor leading to the vault.

**Raxmei** · April 5, 2011, 05:48

Originally posted by Scraper

This raises a question I have about vaults. Is it possible to blink/ teleport etc in to them?

After battling the witch king, who teleported me away, I found myself just below the vault but with no easy access. I would have had to have gone a long way out of my way to get back to the entrance. I didn't want to do that as I had TOed many of the nastier nasties and didn't want to have to face them again.

I decided to try to blink into the vault. No success after many tries. In the end, I dug my way around the vault and close to the original corridor leading to the vault . Cast blink and bam, straight into the corridor leading to the vault.

No, you can't teleport into vaults. If you could then it would be possible for teleports to land you right in the middle of a vault, and that would normally be bad. Forcing teleports to land outside of vaults also makes phase door a more useful escape in vaults since you can rely on it kicking you out of the vault. I believe discussion of changing this is a semi-frequent topic branching off of the difficulty discussion.

Back to the original fellow, as a paladin you could have tried careful application of earthquakes to hopefully open up the entrance.

**myshkin** · April 5, 2011, 06:00

Originally posted by Timo Pietilä

There is a bug in 3.2.0 that doesn't guarantee excellent objects in "8" blocks of GV:s. It is one word fix in generate.c

I decided to look through the logs; this behavior has been in the code since commit 96380c8 (SVN revision 591), from September 29, 2007. d_m fixed it to drop great objects again in git commit 0af51bb on January 22, 2011.

**Derakon** · April 5, 2011, 06:09

Was there any comment in the 2007 commit indicating why it was changed? I couldn't figure out any way to search the checkin history, and besides that was before the move to git.

**myshkin** · April 5, 2011, 06:48

I suspect that the change was inadvertent, but I should not speak for takkaria, whose comment reads, "Refactor the dungeon generation code a bit, so that it reads easier."

I should also note that this change pre-dates 3.0.9a.

**SSK** · April 7, 2011, 23:44

Thanks Timo for also posting about the vault *excellent* item issue on my thread about the "leather shoe store" problem.

In my mind it is clearly a bug and not a feature. As someone else already posted, changing that flag to *FALSE* renders the greater vaults a total waste.

Under the prior system it was always an interesting challenge--PARTICULARLY in the days when artifacts could only be generated once depending on your birth settings. I always felt compelled to try to avoid missing potentially RINGIL or FEANOR or something like that by going for the vault, and it was fun to try to map out a way of tackling it even with a highly unprepared character.

With the flag set to *FALSE*, there is no more risk-reward tension: it's all risk and no reward. Why bother?

So I'm getting Derakon's compiled fix with the flag reset to *true*.

That reminds me:
Derakon--the link you posted on my thread to your compiled version with the fix is broken :-(

**SSK** · April 8, 2011, 00:09

Originally posted by myshkin

I decided to look through the logs; this behavior has been in the code since commit 96380c8 (SVN revision 591), from September 29, 2007. d_m fixed it to drop great objects again in git commit 0af51bb on January 22, 2011.

So guys:

Does this mean that the nightly of 1/22/2011 is the first very recent version that incorporates the fix?

Should I get this nightly or will more recent nightlies be more stable? How much trouble is it to simply get the source of plain old 3.2.0, fix it, and compile it myself (i.e. what tools will I need)? I really hate the useless-vault change and am looking for the best and most stable solution to get back to *excellent* objects in greater vaults...

**Derakon** · April 8, 2011, 01:23

Honestly the current nightly is plenty stable; I'm not aware of any problems with it. If you have access to a compiled version of it it'd be easiest to just use that.

If you want to compile yourself, then you'll need a compiler (gcc on OSX, mingw on Windows). For OSX you can download the Developer Tools installer from Apple's website; it'll include gcc and a bunch of other stuff. Then you just go into the src directory and do "make -f Makefile.osx" (after making whatever changes you want, of course), and after a bit it'll spit out Angband.app one level above you. I'd imagine something similar holds for using mingw on Windows.

Please tell me I'm not the only one to ever do this

Please tell me I'm not the only one to ever do this

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment