Unless I am still misunderstanding the rules, the absence of an 'f' in the hint words does not preclude it from appearing in the solution word.

The rules only specify what must be be there, or rather, what overlaps muct be there. Except for the case when the list conatins a word with 0 (zero) overlap, in which case non of its characters can appear in the solution, but 'f' is not so excluded:

                           offcuts
                          cutoffs     
shortest:3              [sot]       [^hre]
drilling:0              []          [^drilng]    
locked:2                [oc]        [^lked]
messing:1               [s]         [^meing]
irritated:1             [t]         [^iraed]
glory:1                 [o]         [^glry]
modes:2                 [os]        [^mde]
transcribed:3           [tsc]       [^ranibed]
[download]

I think that means that both those are valid?

That my dictionary does contain 'octopus', but doesn't throw it out as a solution is a bug, but I think that the 'ff's are not.

BrowserUk,
Unless I am still misunderstanding the rules, the absence of an 'f' in the hint words does not preclude it from appearing in the solution word.

After people, including yourself, were already working on the challenge - Zaxo asked in the CB if the mystery word could contain letter(s) not present in any of the hint words. I added an assumption.

• There will be no letters in the mystery word that are not covered by the hint words.
Added 2005-01-12 16:10:00 EST

Now the fact that my assumptions changed several times throughout the course of people working on is a reflection of how little progress I had actually made myself. As I indicated in another post - I don't have an official rule book for mystery word puzzles.

To me it was an interesting problem even with my poor assumptions. I have learned from it, so I have upvoted everyone who took time to reply. One lesson is that sometimes that the solutions you dismiss immediately deserve a second look. Another lesson is that no matter how thoroughly you think about a problem, as soon as you open it up to public debate you are going to find different points of view. I have played for a couple of days solving these puzzles by hand and had a method that I thought was smarter/better/faster than dictionary scans so I didn't even pursue it. It turned out coding my manual solution was harder than I thought.

Cheers - L~R

Both my misunderstandings are entirely mine.

It is an interesting problem and one that is hard to specify in words. I just got caught up with trying to solve it using a generated regex and wasn't particularly concerned with keeping an eye on the rules, or even meeting their absolute requirements.

That's pretty much why I didn't publish my code initially. It is also quite a difficult problem to verify your solutions, given the variations in dictionaries etc. The only way I can see of validating is to cross reference with the original (deleted) problem and other peoples generated problems and solutions. I always left the possibility that my code was in error open when I posted questions/solutions.

Adding code to deal with this particular rule isn't a problem, but I have still to yet prevent my code from rejecting possible matches that contain duplicate, required letters. It may be that I won't find a (pure regex) solution to that task, but I am thoroughly enjoying the attempt. I learned a couple of new (to me) regex tricks already.

If my earlier suggestion that you re-write the rules to incorporate some of the separate assumptions into the base description, in an attempt to make the description clearer, came across as 'scolding', it wasn't indended to.

Thankyou for posting the challenge. I am (still) enjoying it. I wish we had more like this on a regular basis--preferably at weekends.

---

Examine what is said, not who speaks.

Silence betokens consent.

Love the truth but pardon error.