Sudtyro2 wrote:Ngisa wrote: Well, I just wanted to shorten the chain:
(5=7)r4c4 - (7)r4c1 = [(7-9)r6c1 =(9-5)r7c1] = (5)r7c5 =>....
Nothing wrong with the full chain...but for a more compacted Eureka format why not write the bracketed section as [(79-5)r67c1] or [(7-95)r67c1]? Analogous to ALS, either term is simply an AHS (Almost Hidden Set).
<flashes the SPD badge>
- Please step out of the car, sir, and put your hands behind your back.
- You have the right to remain silent...
Actually, no need for grammar-policing here
Just something very minor about my other favorite topic: terminology. First, comparing ALS and AHS as two different things is a bit confusing, though it's done all the time and you can't be blamed for that. To me the preferable way to see things is that ALS (almost locked set) is the supertype of both ANS (almost naked set) and AHS (almost hidden set), which means that they're both ALSs. That's the most logical view because they both contain a locked subset strongly linked to a spoiler (or obstacle) which makes them "almost". The only difference is that the ANS spoiler is internal and the AHS spoiler is external to the locked set cells.
Problem is, most people (including myself) use ALS as a synonym for ANS, which leads to ambiguous communication. It's the exact same problem that I have with "Turbot Fish" being both a specific pattern and its own supertype (which is why I rather use "Turbot Crane" for the pattern). Such naming hierarchies are simply bad design, although there are historical reasons why it's gotten that way. (Another source of confusion is that "locked" can also refer to something restricted to a box-line intersection. Thus, something like a "locked pair" or "locked triple" is ambiguous. I really hate it.)
Secondly, what exactly is an AHS? You imply that it's the term (79-5)r67c1. That would seem kind of natural because (79=5)r67c1 would be an ANS in the same cells. However, what makes the latter an ANS is the strong link between the locked set (naked pair 79) and the internal spoiler 5. Similarly, the AHS must include both a locked set (hidden pair 79) and a strongly linked spoiler, but this time the spoiler 7r4c1 isn't in the same cells. Thus the full AHS in this case is: (7)r4c1 = (79)r46c1. The weak link (79-5) is not really part of the AHS or relevant to its definition, although such an internal weak link is only possible with a hidden type of locked set. It would still be an AHS if an external weak link were used:
(7)r4c1 = (79-5)r46c1 = (5)r7c5
(7)r4c1 = (79)r46c1 - (9=25)r7c85
