Hi! > The original idea sprung when writing a document I once called "Scid > for the Impatient". In this kind of tutorial, you introduce the > reader to the basic usage. Scid's on-line tutorial does that, but > with an almost random database [^tutorial]. > > So I think we need a sample database. It should not be too big, just > enough to see how things work. It should be interesting to have for > its own sake.
I agree with you here that a sample is a good thing. Its just I do not know wheterh 25 games or something like that will be enough to show some of scids features. Quite some of them spring off from the statistics, and statistics is usualy the thing dealing with "large numbers". > I like Winter's idea. It's not for Chessbase to decide if > we can use the collection, but Edward Winter. I doubt he > would mind. (And if he does, we would just need to remove > three of the four Tartakover's games to respect > intellectual copyright. That would lose naturalness, > though.) But if we don't want to indulge with it, we > could go with a collection that has no intellectual rights > : The point with intellectual rights is very difficult, in fact. Point here is, that what might be perfetly legal in one country might be perfectly illegal in another one. Its nasty already in germany (as a librarian I've sometimes to do with these issues, not directly but basic knowledge is required) but e.g. there is a huge difference between germany and the US. Therefore, either one does it on his own or gets proper permissions. > Beginning last autumn, I pick a miniature I post in some > chess forum that allows the viewing of games. See [1] for > an example. What I do is simple : upload the weekly TWIC > ; filter the games with a saved search (+2500, less than > 35 moves) ; select the one that appeals the most to me. > > Here is what I think. This collection takes more time, > but I like to do that. It does not prevent one to include > Zurich 1953. It does give some more tactical exercises > (wink wink). I gives some distinction to Scid. It's a > cheap way to have something more to share in some eventual > "resources" page, I think you idea is not bad, though I'd consider not to take just one game but really add the whole tournament and flag the games. Scid offers some flags to notify the user why this one is of particular interest. By this the DB would grow substentially faster and have a good additional value. And there's a very appealing side effect: if I get you right, you do this anyway, therefore it is no additional workload on your desk. I feel this is important here. > [^big-stuff] Building an open opening book and providing > people with a really big database are two very important > ideas. I am having difficulty to find the time for the > Scid's documentation overhaul I began, but it's on its > way. I underestimated the work needed for its information > architecture. I am sure there is problems with the ideas > of database and opening books too. So as for myself, I > suspect I won't deal with those until 2009. These are definitly major projects, this is not the point. Either of it is _very_ ambitious, I think. Both of them could most likely be done by a group of users that reall want to do it. From the work I think it is _easier_ to set up a reference db. As I said one could start with including TWIC, and cleaning it up. (TWIC procudes sometimes doublettes itself.) This, I feel, would provide a solid basis. One could then go back in time work through free sources etc. and add additional stuff, I'd base on tournaments. Always adding a complete turnament at the time with decent and clean headers. This is a stuff that has a lot of time whilst including TWIC would be a bit more pressing as you'd otherwise have about 52 new packages to include just for one year. Actually, for an opening book, things do not balance that well. E.g. one can't just say: ok, we include the Kings gambit and then we have a first milestone. Such a book would be almost unusable. One could start cutting on the depth of variants but I don't think this will be very usefull either. > [^wish] Wish-speak on. The database could also come with > some premiliminary comment, but I don't know how it could > be done in Scid. It is a pity we can provide information > for games, but not for databases. It would certainly be > interesting to have more information about the databases > in the File Finder window. But it seems chess databases > are condemned to remain big bags that are impossible to > maintain strictly. Wish-speak off. Well, this depends. Of course you've all game metadata in your database. You can also store additional metadata about the base itself outside of it. (E.g. STN does this: they provide whole folders of information about the contents of their scientific databases outside of the DBs as such). Additionally, in putting things together one would most likely have the "master db" that is offered for download and some others smaller db's that get merged into the large one if work on them is finished. The most simple approach, far from perfect: my own refdb is a pretty huge one based on Mega2006 + twic + Bundesliga + ... To incorporate TWIC e.g. I collect TWIC not directly into this DB, but I just fetch the issues of one year. I also add other sources of one year into this directory. All this is still PGN at that point. Once a year I create a "yearbook" out of these sources. This yearbook is spellchecked, dupechecked etc. pp. And only once this is done I merge it into my main DB. Scid offers quite some commandline tools to do some things automagically here. BTW: probably a good source of games could be http://www.pgnmentor.com/files.html -- Kind regards, / War is Peace. | Freedom is Slavery. Alexander Wagner | Ignorance is Strength. | | Theory : G. Orwell, "1984" / In practice: USA, since 2001 ------------------------------------------------------------------------- Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php _______________________________________________ Scid-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/scid-users
