[Retros] Some retro active thoughts - database

[andrew buchanan]

Hi Per,

I'd like to respond to your interesting email.

You wrote:

> What would be a better place than the Retro Corner?


Well what about PDB? I have no axe to grind but I have always found PDB to
be a wonderful resource for retro problems (as well as many other genres).

Rather than setting up a new database (and interface) from scratch, the
simplest solution would be to work with the PDB guys to see if it's possible
to extend PDB to give us additional features that we need. They may well
find that these features are not specific to retro, but can be used for
other problems too.

Strengths of PDB:
- Simple, powerful, fast, robust... essentially just one database table for
compositions
- Huge coverage
- Attractive straightforward interface, including dynamic diagrams
- Easy to insert comments
- Multilingual
- Numerous keywords widely supported

Areas where PDB could be expanded (IMHO):
- Need second table to describe tasks/tourneys etc, so that the design
context in which a composition was composed can be understood. At the moment
we are missing this.
- Need to ensure that retro coverage is more nearly complete (in some
sense), without including every sound proof game which can exist.
- Need better quality assurance process, e.g.
	= proof-reading/verification process at the time of data entry.
      = need better process for actioning & closing corrections suggested in
comments 
- Need more work on keywords: more of them, more comprehensive multi-lingual
descriptions, ensure more complete links to compositions etc.
- Need to track changing status of record holders, so it is easy to get
uptodate status info.

Some of these suggestions are technical, some are content-related. What I
would suggest as the next step is we get some kind of  "decision" as to
whether to create a new database, or whether to build on what PDB have
already provided.

If the PDB route is chosen, a retro editor could be appointed for PDB, who
works with the rest of the PDB team to plan the detailed way forward. I am
sure there will be lots of opportunities for individuals to contribute in
building this forwards.

Christmas Holiday Greetings,
Andrew Buchanan.


>Need for a database

>

>The above said gives the thought that we should have some type of database

>for our retro problems. What would be a better place that the Retro Corner?

>It could include first publishing of its type and record achievements;

>there could be as retros, constructional tasks, proofgames, fairy

>conditions etc. Examples of what would be good to have as centralized

>common information:

>

>Historical facts

>-First proofgame with variation in move order

>-First proofgame with single solution 

>-First proofgame with two / three solutions

>-First retro based on 50 moves rule

>-etc

>

>and in the record area

>

>-Longest proofgame from IP 

>-Longest proofgame from A to B. Nobody has tried yet!  

>-Longest proofgame starting from Random Chess (Fischerandom). Nobody has

>tried yet!

>-Promotions

>-etc 

>

>Compare with endgames

>

>Let's look on how well the endgame study community is doing:  the magazine

>EG with primus motor AJ Roycroft publishes copy of all awards in the world.

>There is an electronic collection of endgames by H. van der Heijden, who

>also provides anticipation services. Is there anything that goes unnoticed

>in the endgame world? 

>

>Proofgames and other retros have gained great popularity in recent years.

>The Gert Wilts book from early nineties covered everything in proofgames up

>to then.  But after that we have had an explosion, which is published all

>over the world.  This means that the present results are known to nobody.

>Something should be done.

>

>If we don't want to copy the concept of the endgames, then we could have a

>database gathering all essential that has been done. So far there are only

>part of the new retros reproduced in the Retro Corner. What could prevent

>us from republishing all retro problems that are being published in the

>world (are there any copyright issues?).  Then we would all be up-to-date!

>

>All problem chess achievements will be put into electronic form sooner or

>later; if we do it sooner the task will be easier.

>

>For the moment there is in Retro corner an interesting exercise going on:

>proofgames with the minimum number of moves with a specific feature e.g.

>with a pawn mating on 2nd/7th rank.  Personally I have not been able to

>study all contributions, but I would be really interested to see the end

>results.  But where will these be found?  Answer: by going through all

>contributions there have been? No, that's impossible, the answer is in the

>suggested database.

>

>If the idea is good, then we only need somebody to organize it!   Comments

>are welcome.

>

>Espoo,  Finland  21.12.2006

>

>Per Olin