BBO Discussion Forums: New standalone version of Dealer - BBO Discussion Forums

Alan Davidson announced on that other site that he had prepared a new-version-of-dealer-software which allows more efficient and intuitive predeal specification of the type "spades(north)==2" meaning any 2 cards.

My interest in the project is that it he offers an up to date and maintained Windows executable and may be willing to make the Tricks function (suppressed without explanation by BBO) work too.

Maybe BBO could look into implementing the same predeal and above all restoring Tricks, with or without using AI to replace DDA as in GIBBO.

hi, i don't use the dealer scripts in bidding/teaching tables much.
but i would like to better understand what the gaps are, like what doesn't work.
from your post, i get that it's the tricks function which you would like to use.
can you give an example of a dealer script which you would like to use but doesn't work?
that will help me look into it

 lorserker, on 2026-January-08, 10:14, said:

Great, thanks!

Dealer script has a quirky syntax that could be better documented, but it's still great for simulation. The only huge gap from a skilled programmer point of view is that the tricks function no longer works: it was provided in a separate instance (dealergib1) that simply disappeared when an old server was turned off.
Another gap is that the ability to define vulnerability is not implemented (although that may be true of other implementations also).

From a BBO point of view, it may be a gap that there is no limit on the total CPU usage by a user, only an a cutoff of max 100000 deals per run (IIRC). But I suspect that this is no real issue.

Tricks is used to see how many tricks one will make in a certain denomination, which is often invaluable in simulation.

I posted several scripts that use tricks in the Full Disclosure and Dealer forum a few years ago, but almost all of our posted scripts seem to have disappeared when the forum was renamed to Dealer scripts and the obsolete Disclosure stuff removed.
But as soon as I get to my PC I'll find one and post it here.

Hi @lorserker,
here is an an example I found that looks easy to understand and likely to work.
It's a simple script I wrote to verify my suspicion that the assertion "rightsiding gains little or nothing, even opposite a strong 1NT" was of organic origin.
You can tune it by varying HCP of South or HCP range of North.
Please let me know if it gives problems or you were looking for something different.

Quote

Running this in <https://www.bridgebase.com/tools/dealer/dealer.php> produces the error "Can't open temporary file" because this version is not set up with GiB to perform the double dummy calculation.
It used to work in <http://dealergib1.bridgebase.com/tools/dealer/dealer.php>, barmar mentioned that server being turned off and probably knows how to find any code diffences.

thanks, i understand what you are trying to do.
it's an analysis rather than a spec for what deal should come next in a bidding or teaching table.

for this purpose i recommend that you run the dealer program locally on your computer.
DD analysis on a large number of deals take a lot of resources to compute, and we can't offer it (imagine thousands of people start calling the dealer like that and immediately set off millions of DD solvers)

lorserker, on 2026-January-13, 11:15, said:

But you always did offer it, for at least ten years?
And nobody burned out the server, even with full double dummy rather than AI.
Can you really imagine thousands of people capable of and interested in using the dealer to perform simulations of bridge in an obscure and poorly parsed programming language?
I imagine we are tens, and the odds or more than one at once using the service are small.
But afterwards we post our results and analyse the bridge implications, and that *is* read by thousands, who gain a positive impression of BBO in the progress.

As I suggested, you could set a more effective limit on CPU usage per user if you think it is a real problem.
Or make us pay some small price in BB$ if really necessary to cover CPU costs.

It seems absurd and arbitrary to terminate without real justification a service that was working and added value to the BBO ecosystem. We are not just here to play, but to improve our systems and ultimately, bridge itself. Your competitors can easily offer just a game.

Fred would understand, I suspect also Olivier if asked.