[BillHiggin]

Many conventions have multiple versions/variations. The more complicated the convention is, the more versions there tend to be. Individuals (or partnerships) ought to be able to make their own decisions about which version they prefer (and of course, they are).

The current flat list of conventions will quickly become unwieldy as more versions are added (note there are currently four versions of Bergen). A tree or outline structure might be an improvement.

I see two main types of variation.
1) Alternate incompatible version
2) Extension or tweaks

An example of an extremely variable convention is Two Way Checkback/xyz.
The version included in the BBO-advanced convention card only seems to apply to auctions beginning 1m 1M 1N. I have more commonly seen that referred to as “xyNt”, and even that often includes agreements for 1♣ 1♦ 1N and may include 1♥ 1♠ 1N. My understanding is that “xyz” normally also includes 1m 1y 1M auctions (still with 1♥ 1♠ 1N being only an option). These variations are essentially additive. I.e. they can be divided as:
a} 1m 1M 1N agreements (four unobstructed auctions)
b} 1♣ 1♦ 1N agreements (one unobstructed auction)
c} 1♥ 1♠ 1N agreements (one unobstructed auction)
d} 1m 1♥ 1♠ agreements (two unobstructed auctions)
e} 1♣ 1♦ 1M agreements (two unobstructed auctions)
With the last two being somewhat logically combined (I have not seen a variation including d) without and e) also, but who knows).
An example of alternate agreements with this example involves sequences where checkback is available but not used:
a} Original version used alternate GF agreements
b} More commonly I have seen alternate invitational agreements (and not always the same ones).
Then there are the 2N tweaks:
a} Delayed GF support for opener’s minor (since inv minor denies major
1x 1y; 1z 2N; 3♦ - GF with support for x and lower shortness
1x 1y; 1z 2N; 3♥ - GF with support for x and higher shortness
1x 1y; 1z 2N; 3♠ - serious slam try with support for x and no shortness
1x 1y; 1z 2N; 3N – mild slam try with support for x and no shortness
b} Concentrated Value responder 1 or 2 suited
1x 1y; 1z 2N; 3♦ - GF COV y+♦ or long ♦ (y=♦)
1x 1y; 1z 2N; 3♥ - GF COV y+♥ or long ♥ (y=♥)
1x 1y; 1z 2N; 3♠ - GF COV y+♠ or long ♠ (y=♠)
1x 1y; 1z 2N; 3N – GF COV y+♣
c} Others (I am sure)
It would be nice to define an add-on to implement these which would only actually activate the options chosen above for implemented x, y, z. Normally, the base files would define 1x 1y; 1z 2N as natural invite, 1x 1y; 1z 2♣; 2♦ 2N as some invite with clubs and 1x 1y; 1z 2♣; 2♦ 3♣ as ♣ signoff. In a perfect world, the 2N tweak add-ons would then redefine 1x 1y; 1z 2♣; 2♦ 2N as the natural invite, 1x 1y; 1z 2N; 3♣ pass as the ♣ signoff, and 1x 1y; 1z 2♣; 2♦ 3♣ as whatever 1x 1y; 1z 2♣; 2♦ 2N was in the base file (ok – I am dreaming).

Note that if all the possible variations were included as single files that the list becomes enormous (especially since I only listed some of the variants).