[david_c]

Seems to me that Blackwood is a problem because it can be applied in so many different sequences. So the file will contain some instances of Blackwood being defined, but we can't hope to list all the sequences where it applies. So from the users' point of view, some of their Blackwood bids are automatically alerted, but others aren't. Is this going to cause problems?

[Greyboots]

Blackwood (or RKCB) sequences illustrate the general problem with data entry, data validation and data update with the program/data file structure as it is. Enumerating all possible sequences in which RKCB can be called upon will be a Herculean task. Let alone then entering all the RKCB responses for all these sequences. After that you have the bigger problem of testing all these sequences. Finally, after having entered and validated all this data, if you now want to change the RKCB responses from 30-41 to 41-30, you have the equally daunting task of finding all the RKCB sequences and then modifing all these sequences.

Maybe the program/data structure could be updated to allow you to define the RKCB responses, and then define a set of rules/states when these RKCB responses should be utilised. Ideally the RKCB responses should be defined as either absolute (e.g. 5C always shows 30) or relative (e.g. next suit up) shows 30. This would enable the RKCB responses to slide in anywhere, rather than always be fixed to 4NT/5NT etc.

This approach could also be used in the earlier rounds of the auction, e.g. the responses after 1x (X) could be defined once, and then invoked 4 times. e.g. 2NT (absolute) = good raise, double jump (relative) = pre-emptive, etc. .