Well, now I am wondering if it wouldn't be possible-- theory about UUIDs notwithstanding-- to come up with procedures for group contribution to ensure higher quality? So, instead of adding willy-nilly, try to prevent the "garbage in?"
c On Sun, Jan 2, 2011 at 1:14 PM, Gerd Lorscheid <gerd.lorsch...@onlinehome.de> wrote: > Hi, > > you still have the problem, that two users on two systems enter the same > game. Now these two games are duplicates viewed from their internal database > ids, but not for me. Unique IDs in databases are mainly useful to allow > quick referencing inside of the database. What we would need and what does > not exist is known by databases as natural keys. I guess a composition of > them is what you are looking for. But there is no subset of attributes, > which allows to determine identity of two games. That is why only a smart > duplicate detection can help. > That is why for example discovery of a computer network with all its > components is such a nice job. Very similar to our problem. > > Gerd > > > > -----Ursprüngliche Nachricht----- > Von: Benoit St-Pierre [mailto:benbon...@gmail.com] > Gesendet: Sonntag, 2. Januar 2011 22:55 > An: Scid Users List > Betreff: Re: [Scid-users] ScidBase? > >> unique IDs to ensure a *high-quality* cannot be achieved. How should the > ID be calculated? > > The ID is simply given and maintained by the database facility. In > relational database theory, I believe that to have unique IDs is > axiomatic. I am not sure that it's possible to have relational > databases that work without unique IDs. > > This ID would guarantee the internal consistency of the database. To > have something like external consistency, we sould need a service akin > to URI or DOI conventions. > > The way I envision it, we could completely kill off duplicates by > first insuring that we have the proper metadata and the correct > gamescore. Then, as soon as a game is sufficiently similar to thess > corrected data, we simply would delete this game and replace it with > the sanitized data. > > That hundreds of thousands chessplayers hand-pick and correct chess > scores and chess metadata by hand is beyond me. It also runs against > every principles on which is based archiving. > > In fact, that we can't yet download a sanitized database for Scid is > still beyond me. > > More on that has been already told. Search for CentriScid in the archives. > > *** > > That said, I know that I am speaking from a theorical standpoint. I > absolutely have no idea if that's possible in practice, considering > the meager resources we have for now and the way to "hunt and gather" > chess games is deeply rooted in chess culture. So please bear in mind > that I am in no way asking anyone to do anithing here. > > ---------------------------------------------------------------------------- > -- > Learn how Oracle Real Application Clusters (RAC) One Node allows customers > to consolidate database storage, standardize their database environment, > and, > should the need arise, upgrade to a full multi-node Oracle RAC database > without downtime or disruption > http://p.sf.net/sfu/oracle-sfdevnl > _______________________________________________ > Scid-users mailing list > Scid-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/scid-users > > > ------------------------------------------------------------------------------ > Learn how Oracle Real Application Clusters (RAC) One Node allows customers > to consolidate database storage, standardize their database environment, and, > should the need arise, upgrade to a full multi-node Oracle RAC database > without downtime or disruption > http://p.sf.net/sfu/oracle-sfdevnl > _______________________________________________ > Scid-users mailing list > Scid-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/scid-users > ------------------------------------------------------------------------------ Learn how Oracle Real Application Clusters (RAC) One Node allows customers to consolidate database storage, standardize their database environment, and, should the need arise, upgrade to a full multi-node Oracle RAC database without downtime or disruption http://p.sf.net/sfu/oracle-sfdevnl _______________________________________________ Scid-users mailing list Scid-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/scid-users
