I have collected a pretty big amount of (free) PGNs that I organized with my (free) chessDB database into 5 categories using the ECO codes: A,...E

I then subdivided into slightly more manageable (less big) files like:

BenoniA60-79.pgn [29 MB], DutchA80-99 [62 MB] and some catch-all like A00-59.pgn [569 MB]...

These PGNs can be manipulated by a program like Chesspad (free) or imported into databases.

a) I am more than willing to share them and despite the size, since we meet regularly, sharing should be possible. (With time I am also planning to reorganize them so that sizes will be more manageable.)

b) Could we possibly start a common project where those files could be augmented, improved and organized so that our little group would have an impressive common database that would help us improve the general level of opening knowledge? 

c) Anybody has ideas about how to make that project manageable?

Raymond: I don't have a lot of experience using PGNs, and would be more than willing to share any I find.  Would this be more than the 500,000 games on Chessmaster or the games available for free from Chessbase light?  Also, are commercial programs too expensive? 

If you have time, could you also explain how one is supposed to use the PGNs to study openings?  Is there some kind of systematic approach other than to just studying individual games of masters and grandmasters that are using a particular opening? 

The advantages of PGNs over the databases like Chessmaster's are

1) PGN is a standard that most databases can read

2) since it's a standard it's ideal for sharing

3) Most playing programs can output the games in PGN

4) It's easy to read and modify with a text editing program

5) You can find a lot of them "out there"

6) There are free database programs like chesspad which act directly on PGNS

7) The big database programs like chessbase (300$) and chessdb (free) will store games using their own format but can read PGNS and translate their own format into PGN

PGN databases can not really replace books where strategic ideas are explained but they should be used as complements since they can be updated easily with recent games, your own games and your own analysis. (I can show you on Thursday) Buying chessbase is of course simpler but I am hoping that with some cooperation we can develop a database that would be equivalent in power, contain local games and would be affordable for all. 

I'm interested in this.

So far, I've just used the database that came with Chessmaster and occasionally browsed www.chessgames.com to look for additional info, and picked up games from the large archive at:  ftp://ftp.pitt.edu/group/student-activities/chess/PGN/

I'd be curious how functional chessdb is compared with chessbase once you start to explore databases beyond that included in chessmaster. While a database of many millions of games sounds great, how much incremental value are you gettting if you only play a handful of openings on both sides of the board?

From another perspective, while $300 is indeed $300, how much time would we spend building something comparable using chessdb.

As a final comment, I run a couple of web sites for sports clubs I involved with, and have plenty of bandwidth available should we want to have files available on the web for download.

Richard,

Let me try to cover all your points...

I am pretty happy with ChessDb as long as I split the databases in manageable chunks. E.g. A "Sicilian" database with a million games is too slow for my taste and laptop power, while a "Dragon" or "Najdorf" with 1 or 2 hundred thousands is fine. (Splitting as the drawback that some fringe variations will force you to change database, but with the ECO codes -and experience- as a guidance, it's OK. Also, ChessDb allows you to keep several databases open at the same time so navigating from one to another is pretty painless.)

It is true that buying ChessBase or ChessAssistant would be more practical for many people; however my goal here -thinking more as a chess instructor than a chess player- would be to make the databases available to a "community" of players. This brings in the following points:

1) They may not all have enough interest  in Chess to spend the money.

2) If we want to establish a "standard" to share games, the combination of freely available PGN games + ChessDb, is attractive. (If 10 players share databases, the total investment cost for the group is still $0 and not $3000)  

 3) There could be in my mind a majority of passive users - with no interest in contributing games- and those could use a tool like Chesspad that handles PGN files directly. (It can get real slow when the database is big but is easy to understand and use.)

4) The active contributors could use any database as long as it is able to import from and export to PGN format. (!There may be some copyright issues with commercially available "commented" games.)

5) I think this could work as long as the role of "active contributor" is not too burdensome. (Altruism is not necessarily a strong long-term motivator.) I suggest that the contributors could restrict themselves to openings they have an interest in.

6) There would obviously be some work to set up the system (Sharing mechanism?How do we split the databases? Guidelines for contributors...) and this could only be done by a little group.

7) I am of course also not sure that it can work.  I have my own questions...Is there any interest out there? Are there enough people wanting to contribute? Does anybody  know if the same type of idea is already implemented somewhere? Would databases of local games or augmented by local games be attractive?

Thank you of course for your offer of download bandwidth!