Hello and merry Christmas! First I would like to thank all who have spend time developing gnubg and keeping it working. The first time I have used gnubg was around 2003 and it’s great that it is still around!
I see that development has slowed down and I think it’s quite normal that after 20 years priorities shift for contributors. So I think it’s also important to attract new contributors. There are now much more software developers on this planet than 15, 20 years ago and interest in AI is very high. I think it’s possible to find new people. I see some obstacles currently: 1. Build issues Last weekend I tried to compile gnubg on my M1 Mac and I couldn’t make it compile under macOS. Running it in a virtual machine with Linux, it was compiling however. Autotools is quite difficult, at least for me. 2. Separation of concerns If you want to play around with inputs and evaluation, develop a bot or another GUI it would be very helpful to have a core gnubg library without application code. For example a simple C or Python function with match and position as input and equities as output. This is currently missing or I can’t find it. There have been people who ported gnubg to Android and web. Those initiatives are not ongoing, code changes have not found its way back to the gnubg repository. https://github.com/alcacoop/libgnubg-android https://github.com/hwatheod/gnubg-web 3. Repository: Ease of use People with less than 15 years of development experience have probably never used CVS. Everything is git today. The most visibility gnubg would have on GitHub.Some developers care about their contribution statistics on GitHub; commits on Savannah are basically hidden. GitHub also provides CI (Linux, Windows, Mac) free of cost for public repositories. So my suggestion is to first move the repository to GitHub. I’m willing to help with the move and could also create GitHub actions for CI that automatically build gnubg with every pull request. What are your thoughts? Best, Carsten
