Well, in the Qg4 lines I think Black should castle. White's kingside attack looks more dangerous than it really is, sooner or later they will retreat their queen and thanks to their weak pawns and also because they can't really prevent Black from advancing their e-pawn at some point White will have a worse game and can only hope for a draw.
White plans with an early ...a4 (aiming to play their darksquare bishop to a3) has the major drawback for them that Black can typically win the a4-pawn (usually Bd7 x a4, with Black's queen on a5 which protects the bishop). The pin of the bishop by the a1-rook doesn't help White, because Black can always protect the bishop with ...b5. Also, the bishop will then permanently attack the c2-pawn. So I think the ...a4-variation basically blunders a pawn for White and they will not get sufficient compensation.
Most interesting but also most complicated to play are the Nf3-variations (but when White does not transpose to an ...a4-system). Typically Black will castle to the queenside (after completely blocking the center by playing ...c4), and White's semi-open b-File is not worth much because thanks to their doubled pawns on c2+c3 White's wings are badly coordinated, and Black can simply defend b7 with their 'French' bishop. So until White is able to generate serious threats against Black's king, Black's counter-play against White's king (who is usually castled short) will have grown at least equally dangerous. Very, very interesting variation with good chances for both sides to make a full point.
Because I'm an 1.e4 player, too, I also need to know something against Bb4. What seems to work best for me as White is the system 4.Nge2. It's actually a gambit (dxe4), and IMO White gets great attacking chances if Black tries to defend their e4-extra-pawn by ...f5. But that's what most Black players will do, because only very few Black players (on our level!) know how to defend against Alekhine's gambit correctly. If they know, then White will simply regain the pawn (Nxe4), and keep the longterm advantage of the bishop-pair in a position which has more an open character (compared to the typical French positions), and Black has a very hard time to get an equal game.
Most other gambits or pseudo-gambits against the Winawer (e.g. 4.a3 etc.) I think are not so dangerous for Black as the 4.Nge2.
Well, of course the good old Poisoned Pawn Variation can still be played instead of castling (against Qg4), after all, great players like Botvinnik and Uhlmann have used it, so it sure offers Black good chances. I don't like it so much, because the amount of theory behind this variation is so enormous (compared to 7...0-0), and not knowing the theory very well in this line is deadly (for both players). 7...0-0 I consider to be the 'safer' option.
greetings
Just wondering what people out there thought were interesting ideas related to the black handling of the Winawer variation of the French defence.
Man that is a gorgeous sunset! @8:48 PM in Winnipeg.
Jordan