Thanks for dropping in. I am glad to have helped with 11ac. On 3dn we will simply have to agree to differ on our preferences for this type of clue.

Thanks for the blog; thanks also to Thomas99, crypticsue and Pelham Barton for your comments.

As regards 11a, Pelham Barton has interpreted my intentions; I don’t mind the odd linking word (“but” in 22a also) if it assists the surface – something I know you have mentioned before Shuchi.

As regards 3d, Thomas99 has put it far more eloquently than I ever could!

As regards anagram indicators, I do my best to steer away from the more obvious ones, and in my quest for originality, I will scour my Chambers Thesaurus, and cross-reference what I may find with the Chambers dictionary. In the instances here that have been commented upon:
Another – “different, alternative, not the same, variant”
Suit – amongst other things, “become” (I probably won’t be using this one again)

Finally, I’m disappearing shortly on an Antipodean adventure, so may I also use this opportunity to thank Gaufrid and his team of bloggers, all those who have been kind enough to comment on my puzzles during the year, and my compiling colleagues at the FT and elsewhere, a very Happy Christmas, and best wishes for peaceful and successful 2012.

I agree that posthaste is meant to be split. I don’t think it is all that unlikely that someone might realise this without being taught it as a convention. Surely Hamilton is more likely to be trying to trick experienced solvers who have got used to certain other conventions and consequently don’t think of looking at the words-within-words separately? If posthaste can also be a synonym of despatch then I suppose it’s also a sort of &lit. I wouldn’t be too hard on him here anyway, especially as posthaste was presumably a two-word phrase which at some point (possibly after a hyphenated spell?) got turned into a portmanteau word. I’d agree that some sort of indicator or at least a nudge like a question mark would make it more above board (above-board? aboveboard?) though.

11ac: I think “may” is a linking word and “group” indicates the hidden part.

3dn: This could be taken as a one word straight definition – after all the crossword is not actually billed as cryptic. However, it is presumably based (as shuchi suggests in the blog) on an unsignalled requirement to split a clue word. Fine if you regard crossword solving as a series of conventions to be learned, but I will always regard it as unsatisfactory.

10dn: In the blog, shuchi asks about “another” as an anagram indicator. I cannot find a justification for this. For me, the same goes for “suited” in 14ac. Can anyone give a convincing justification for either of these?

Post updated.

20a The chef is Roux and the fat is lard.

Fun puzzle, but wasn’t familars with NW by W in that order! Thanks for the blog.