Macintosh Chess Database

Well, then what's your complaint as Stockfish runs on Mac natively.

I never complained about the strength of an engine. I simply want a DATABASE for Macintosh.

oops sorry, confused you with Mark100net.

Ok, sorry. Now that that confusion is out of the way, any suggestions for a name?

I don't care what programs are named, I care about how well they work :)

I am developer of WinBoard / XBoard, and always interested in ideas for improving it. I am not an active Chess player, never use databases, and have never seen ChessBase, though. But it seems to me that starting from an existing open-source program like XBoard or SCID, and modifying it to do exactly what you want would be the logical thing to do, as it would save you about 99% of the work. Possible downside is that it could not be used if you want to keep the sources secret: the GPL forces you to make your modified version public under GPL as well.

XBoard is available for Mac, and currently supports only very basic database functionality: it can load databases in PGN format (huge and slow...), and you can search positions in those based on 6 different criteria:

• Exact position match
• Given position with extra material on the empty squares
• Exact match for the Pawn structure with same piece material anywhere
• Exactly the same material anywhere on the board
• Material within a given range with a specified material difference
• Material within a given range

where it optionally supports also matching color-reversed or horizontally-flipped positions.

That could be very primitive to what you want, but adding new search modes is basically a job of an hour, or so, where writing a program from scratch can easily take several men-year.

I agree. I think the number of open-source programs currently available will save us a ton of time. However, we should note the reasons why other databases like ChessBase are more popular, and develop ours accordingly.

The main reason ChessBase is my preferred database is because it has a clean intuitive GUI (I wish the mac would adopt the ribbon in more programs) and much more importantly has a wealth of well supported addons, training software, and extremely prompt automatic database updates.  The package as a whole is far more valuable than any of the individual aspects.  Running in coherence mode on my mac air and mac pro it is indistiguishable from a native mac app for me.  The only problem I have had is to remember to control-v copy from ChessBase and then command-v paste into another program.  While I would like to have a mac version of the ChessBase ecosystem as a whole, it is not that important to me to have a mac database without the rest of the ChessBase support system.  SCID is fine for that for my needs or one of the other freebie database is fine for that. 

Ok, I think at our current stage of development, our next job is to contact some top chess professionals and developers to see if they would be interested in helping out with the database.

It would be wise to decouple the projects of creating a database application to manage the data on one hand, and collecting the data and compile them into an actual database on the other hand. The second task is not platform dependent at all, and it would be a waste of effort to just do it for Mac.

Making a database and keeping it up to date is a business that could occupy you for the rest of your life.

Any ideas on the overall style? If anyone knows about the ICC for Mac design, I considered something of that caliber, but perhaps without some of the uglier eccentricities.

Does that mean you plan to create it from scratch after all? Existing interfaces of course already have existing (though often highly configurable) styles.

The interface, yes, will be mostly, if not entirely from scratch. The database and the engine, on the other hand, will not.

You plan on building an engine rather than simply providing UCI plugin capability?

Not building, but using. Stockfish is open-source :)

Why would you incorporate an engine into the database rather than simply providing a universal interface for engines? The whole point of the UCI API is that engines are engines, databases are databases and so on. You don't need to combine them, you just need to use existing standards correctly -- and that code should be something you can take from XBoard or other open source chess playing programs.

Ok, it will be able to incorporate other engines, but Stockfish will be automatically included. I think a lot of people like the idea of a "package deal."

OK, good luck then. I am only interested in the database functionality, as the interface already works.

HGMuller, I found this looking around: http://chessx.sourceforge.net/

Looks like a project is already underway in this area with some success. 

The ICC interface for mac is the paragon of bad design.  It has a lot of smooth colors and prettiness that really inhibit the functionality.  It is completly non-intuitive to tell which functions have tabs, which things have menu items and where certain actions have output.  Horrible. It is the one case I have experienced where I run parallels and use the ICC windows program rather than the mac because it so much better designed and so much faster.