Hi, and welcome to this site! My name is Bo Haglund and I live in Stockholm.

My web site contains download information for my bridge double dummy solver DDS jointly with Soren Hein and for my Sudoku programs.

DDS is included in a development repository, bridge-dds, for DDS on GitHub for interested contributors. Many thanks to Brian Dickens for his great effort to help putting DDS/DDD on GitHub! Thanks to Pierre Cossard who contributed the code for multi-threading on the Mac using GDS. My website remains as the official place for publishing DDS distributions.

2018-05-30:  DDS 2.9.0 beta is released:
DDS 2.9.0 beta has a new, improved software architecture. In addition to the current multithread alternatives WinAPI, OpenMP and GCD, a number of other alternatives are available: Boost, STL and TBB threading. Other multithread alternatives can be added, as required. Adding a new multithread solution is easier in DDS 2.9.0 than in previous releases.

For multithreading, DDS 2.9.0 autoconfigures an optimal mix of threads using the full memory transposition table, and threads using the small memory transposition table, taking into account the number of processor cores and memory available. This often leads to a faster solution than if either the full memory or the small memory solution is used, as is done in the previous DDS release.

In DDS 2.9.0, it is possible to have multiple threading systems compiled in the same library. With a user callable function, SetThreading, you can select the one you want, in case there is more than one.

All documents, source code and DLLs for DDS 2.9.0 beta are obtained here. The DDS 2.9.0 development project is situated at GitHub.
There you can find more information on what is new. You are welcome to try out the new DDS! Please give us feed-back, we need that for making 2.9.0 a real release.

2016-03-21:  DDS 2.8.4 is released:
DDS 2.8.4 includes a lower-memory DLL version that require much less memory. E.g. for a single thread on a PC, memory usage is about 24 MB rather than 110 MB. For 12 threads, the numbers are about 89 MB rather than 1.1 GB. The lower-memory DLL is however about 11 - 14% slower.

The SolveAllChunks functions are now aliases to SolveAllBoards. For details, see the release notes.