From NetHackWiki
Jump to navigation Jump to search


It appears to be possible to {E}ngrave 'Elbereth Elbereth' in 1 turn using a soft gemstone in vanilla. This is in conflict with the description on the main page and if corroborated, should be updated to reflect true behavior.

This is my first comment on this wiki, so excuse formatting issues if I screw it up. --Kaeroku (talk) 22:39, 3 October 2012 (UTC)

Soft gemstones will write in the dust. Hard gemstones will cut into the stone, making semi-permanent engravings. It should say something like "You write in the dust with the <x>". See the quality section on the page. -- Qazmlpok (talk) 12:07, 4 October 2012 (UTC)
Yes, I'm referring to speed though, not what they cut into. See the Speed section on the page. If you are in-game, it's very easy to test that the information contained in the Speed section is incorrect. I'm not sure when it was last updated or whether it was true for prior versions, but it certainly is not, now.--Kaeroku (talk) 03:18, 5 October 2012 (UTC)
It's already reverted by somebody else, but the edit you've made to the main page saying hard gems make fast engravings is wrong. You can test this by going into wizard mode, wishing for and engraving "Elbereth Elbereth" with a diamond (which is a hard gem, versus soft gems like worthless glass or many other valuable gems), and looking at the turn count before and after. --Tungtn (talk) 04:22, 5 October 2012 (UTC)
Also, speed depends on what it cuts into - or rather, on whether it cuts into it or just writes in the dust, which again depends on the type of object. The article does actually point out that engraving with soft gems is fast in the Quality section. Hard gems, however, engrave slowly. See also the references I added. (I might add some more later, too.) --Bcode (talk) 04:29, 5 October 2012 (UTC)


It appears to be possible to (E)ngrave 'Elbereth Elbereth' in 1 turn using a noncursed athame or Magicbane. Dunno why, though. Bug? --Renx 10:16, 14 April 2008 (UTC)

Yeah, according to the code the duration of your engraving is (chars/10) , which is rounded down due to C behaviour. So you can basically go for 19 chars per turn with any fast method. --Renx 10:20, 14 April 2008 (UTC)
Edited the article to reflect this finding. --Renx 10:26, 14 April 2008 (UTC)


The #name command is certainly related to engraving, even if only partially. It certainly rises to the level of deserving a mention on this page. An in-game message speaks of your hand slipping while engraving the #name on the object if you are exploiting the artifact naming bug. It is difficult to understand the behavior of object names in Nethack outside of making some kind of distinguishing mark on the object itself. The mechanics are entirely different from the engrave command, but that's Nethack.

(comment left by Ckbryant on 15:34, 6 June 2008 according to talk page history --Rogerb-on-NAO 21:54, 6 June 2008 (UTC))

Please sign contributions on talk pages.
It's only related in that they are both forms of internally stored writing and the message mentioned happens to use the word "engrave". The current wording on the page implies there is some underlying link in game mechanics. As you note, there is not. Since the message mentions engraving however, it is reasonable to mention it on this page. I suggest the wording of the note is changed to:
  • When you name an item, it is possible to get a message such as "While engraving your hand slips." The mechanics of engraving[1] and naming[2] are separate and quite different; the message is part of an internal mechanism to preserve the unique naming of artifacts. This is certainly a special kind of engraving, because it is absolutely permanent and requires no tool at all!
Unless there are objections, I will do so shortly.
--Rogerb-on-NAO 21:54, 6 June 2008 (UTC)
Suggested message edited --Rogerb-on-NAO 10:16, 7 June 2008 (UTC)
That certainly seems like a good compromise to me. I happily support it.--Ckbryant 16:56, 12 June 2008 (UTC)
Cool. Done. --Rogerb-on-NAO 10:26, 13 June 2008 (UTC)

Charged Wands

Anyone know how many charges are used up for wands? Also, is there some limit on how many letters you can engrave, or is it a cost per number of letters? AileTheAlien 03:21, 13 January 2011 (UTC)

In vanilla, it should cost one charge regardless of how many letters you write. NetHack Brass however requires one charge per letter plus one charge overhead plus one charge if it's a wand of fire (or lightning presumably. Getting this info from Elbereth#NetHack brass) -- Qazmlpok 03:38, 13 January 2011 (UTC)



The text currently says "Riding also makes it more difficult to engrave properly", however, I can't find the string "mount" anywhere in engrave.c. What is the basis for that statement? Derekt75 (talk) 15:41, 26 July 2013 (UTC)




    	return (boolean)(!u.uswallow &&

    #ifdef STEED

    			/* Restricted/unskilled riders can't reach the floor */

    			!(u.usteed && P_SKILL(P_RIDING) < P_BASIC) &&


    			 (!Levitation ||

    			  Is_airlevel(& || Is_waterlevel(&;


    #endif /* OVLB */

--Ozymandias (talk) 15:50, 26 July 2013 (UTC)

but that doesn't make it "more difficult to engrave properly", that makes it "impossible to attempt to engrave", right? Since the line was in the spot about the chance of an Elbereth working, the line implies that someone with Basic riding skill will wind up with a smudged Elbereth more often when riding than when walking. but I don't see this in the code. Should I delete the sentence from the article? Derekt75 (talk) 18:48, 26 July 2013 (UTC)

I was curious and checked it out. It is much, much harder to engrave when riding because your steed will walk all over the engraving. The actual erosion call is at monmove.c:336. Steeds erode engravings on each of their turns, one guaranteed letter per turn. (and *not* upon movement: steeds effectively rest each turn, at monmove.c:512). Thus, if you're at normal speed sitting on a warhorse (speed 24) and try to Elbereth in the dust, two letters are guaranteed to erode, and it is thus impossible to Elbereth in one turn. You can successfully Elbereth only if there is already an engraving, and the two erased letters happen to be from the old engraving. Even then then chances are low (only reaches 50% after 6 Elbereth attempts in the warhorse scenario), and it will be quickly wiped out in subsequent turns anyway. Gonr (talk) 22:40, 14 February 2014 (UTC)

Chance of successfully engraving "Elbereth" with fingers

This article states that every letter you write in the dust with your fingers has a 1/25 chance of being misengraved. It then claims that the chance of succesfully engraving "Elbereth" with your fingers is 72.6%. The actual chance of successfully engraving a message with your fingers is represented by .96n, where n represents the number of letters in the message (.96 = 24/25). Elbereth has eight letters, and .968≈72.1. So, unless I'm missing a variable, you have about a 72.1% chance of successfully engraving "Elbereth" with your fingers on your first try, and I think the article should be changed to reflect that. Mykal (talk) 15:50, 31 July 2014 (UTC)

When you misengrave a letter, there's a 1/94 chance that the random ASCII character chosen to replace it (engrave.c, line 1002) happens to match the one you were trying to write (and, since Elbereth is case-insensitive, a 2/94 chance that the misengraving will have no practical effect). Thus, the *true* probability of each letter changing (other than simply flipping case) is 1/25 × 92/94 ≈ 0.03915 rather than 1/25 = 0.04. Accounting for this, the probability of writing a working Elbereth is indeed about 0.72652. --Ilmari Karonen (talk) 01:53, 1 August 2014 (UTC)
OIC. Sorry. Mykal (talk) 19:35, 1 August 2014 (UTC)

Table for erosion of semi-permanent/permanent

Something is written here in the dust. You read: "Elbenerf".

With the changes to Elbereth in 3.6.0, I have become a lot more interested in the behavior of semi-permanent engravings, and I didn't see any detailed information (i.e. beyond "0 or 1 characters") in the usual places. So I dove into the sources, and not only found answers, but discovered the information was even more important than I thought -- the exact same behavior also applies to the magical degradation of "permanent" engravings, although there's only a 50% chance of it happening. So I've compiled a table and put it here -- I hope it proves helpful.

I have several thoughts about deficiencies or possibilities to improve it; hope someone can take a look at these:

  • The headings for the first two columns are far too long. If anyone can come up with shorter wordings, please please fix it!
  • Related: the "Semi-permanent" should span columns 2 and 3, if possible, with "chance of..." and "percent" as sub-headings. I'm not sure if this can be done, and if so how to do it.
  • Should there be another column, showing the effective percentage against permanent engravings? This would be half the "Percent" column, for magic cases, and 0 for the others. I'm not sure whether this is helpful or needless clutter.
  • Is the column with probabilities as ratios (i.e. 1 in 26 and such) helpful? Clearly for the dn cases it's not, but I felt it was helpful for the first 3 entries, and not very harmful for the rest.

-- 05:06, 5 March 2016 (UTC)