Michael Ayton wrote on 06/21/09 at 22:51:19:
Some time ago now, there was a very interesting Flank Openings thread concerning whether, after 1 c4 c5 2 Nf3 Nf6, 3 g3 or 3 Nc3 was the more 'logical' way for White to negotiate the minefield of transpositions to reach the positions he wanted to play.
One line discussed was 3 Nc3 Nc6 4 g3 d5 5 d4 e6 6 cd. I was surprised to find that NCO mentions only 6 ...ed here, transposing to a Tarrasch. But what should happen after 6 ...Nd5? OK, 7 Bg2 Be7 is a Keres-Parma which White is probably angling for, but Black has other 7th moves -- 7 ...cd, or even 7 ...Nc3. Can anyone tell me the practical/theoretical status of these relative to 7 ...Be7? After 7 ...cd, for instance, play usually reaches a (presumably equal) ending, but is this just tediously equal, or does Black have to be really careful? And has either side any other options?
3. Nc3 with 4. g3 is better.
This is covered in OFWAK 2.
There is a lot of analysis (way too much for me to want to write it out).
The game Gulko-Vaganian, Erevan 1999 covers 7...Nxc3
Kramnik-Lautier, Horgen 1995 covers one branch of 7...cxd4, Kramnik-Lautier, Belgrade 1995 for the other, but an improvement is given for white on move 19.
In a recent game in the latter line played, I believe, by Jan Werle, black is basically losing in this line, as Werle found an even better improvement.
These lines are all considered +=, not equal.