MatPlus.Net

 Website founded by
Milan Velimirović
in 2006

21:04 UTC
ISC 2024
 
  Forum*
 
 
 
 

Username:

Password:

Remember me

 
Forgot your
password?
Click here!
SIGN IN
to create your account if you don't already have one.
CHESS
SOLVING

Tournaments
Rating lists
1-Oct-2024

B P C F





 
 
MatPlus.Net Forum Promenade "Problems by ..." - worthsome project?
 
You can only view this page!
(1) Posted by Hauke Reddmann [Friday, Sep 13, 2024 11:21]

"Problems by ..." - worthsome project?


As chance goes, Walter Szameitat had the same position as I now have -
1st chairman of the chess club of Wilhelmsburg. That explains how I came
to problem chess in the first place (also, I was always attracted to teh lulz),
and why on my desk are his (still in stamped form) problem notices.

I already ran a retrospective in my club zine; editing a booklet
"All problems of Walter Szameitat" (assuming his records are complete)
would cost me at most a month (he was much less prolific than me,
maybe <100 problems). Neither the financial aspect would drive me bankrupt.
Most annoying part might be everything is written in Sütterlin font
(gackt - but in principle I can read it, I'm a man of many wonders :-)

But where would be the demand? Do you think more than 3 hardcore chess
historians would be interested? He wasn't a "big shot" either.

Anyway, someone with access to WinChloe might tell me how many problems
by him are recorded. Schwalbe and YACPDB both have ~10.
 
(Read Only)pid=26190
(2) Posted by Kostas Prentos [Friday, Sep 13, 2024 14:46]

WinChloe has only 15 problems by W.S. All are orthodox mates in 2 and 3 moves. He won 3 prizes and 1 honorable mention.
 
 
(Read Only)pid=26191
(3) Posted by Michal Dragoun [Friday, Sep 13, 2024 20:45]

I am not hardcore chess historian, but I would buy collected works of any composer. And my experience from time to time editing of collected problems of some composers in „Gallery of Bohemian composers“: almost every composer had rather detailed evidence of his works. Almost all this evidence is lost or become lost. So having the complete output of some composer is a good oportunity to save it in some way. I would prefer - as a traditionalist - printed booklet, but definitely it is not the only option.
 
   
(Read Only)pid=26192
(4) Posted by Kevin Begley [Sunday, Sep 15, 2024 12:55]

Problem databases have the capacity to provide everyone with a complete record of every composer.
Booklets which sit behind a paywall mainly serve the seller (Amazon).
They are a disservice to the problem community, because they create a financial disincentive to complete databases.

Better to fill in the database, and provide everyone access.
That leaves room for booklets (behind paywalls or otherwise) which examine problems at a greater depth.

Tangentially (but hopefully not too far off topic)...
Our databases need some remedy.

1) they should provide notable problems (as you will find in the best games database).

Sorting problems by awards, and by republications, can be helpful to locate a composer's most notable problems, but these features are insufficient.
The author might have some say in the matter, but ultimately the problem community (as much as it exists around a given database) should decide by some voting mechanism.


2) Problem databases should add some basic biographical information, for composers (Composing Titles, Judge status, Solving titles, OTB titles, Correspondence titles, DOB, DOD, current/previous nationality, birthplace, photo, FIDE ID, etc).

They also need some space to appreciate composers who are notable outside of the problem chess arena (such information is beneficial in advancing our artform).
I would also argue some fusion with a games database would be wise.


3) They need expanded fields for journals which publish problems (or tournaments where problems are published).

Country of origin, period it existed, chief editors, contributing editors, name changes, image(s) of their logo, etc.


4) Databases need a methodology to share entered problems (at least where all their terms agree).

Presently, every new problem entry must be tailored for each specific database.
People who want to contribute to completing problem databases are discouraged, because their altruism benefits only a specific database (not all databases).
An agreed methodology to contribute to every database, simultaneously, would improve all databases (allowing database developer competition to shift from problem sets advantages to feature advantages).

An extended PGN -- for adding/storing problems to a problem database -- would allow anyone to cultivate their own, specialized databases (and booklets like the one suggested in the original post would simply write themselves).


The best thing about problem chess: this is a hobby which requires the smallest financial investment, where any talented devotee can make fairly significant contributions.
That is not the case for almost any other hobby.

You needn't purchase special chess composer shoes, clubs, rackets.
You needn't rent a lighted field.
Rarely do you need to purchase special software, or a library of booklets (many of which you'll rarely open).

The more we add financial burdens (note: I do not deny there is value in financial incentives, even in this artform), the more we negatively impact our artform.
 
   
(Read Only)pid=26195
(5) Posted by Hauke Reddmann [Sunday, Sep 15, 2024 18:23]

@Kevin:

0. It will be far cheaper for me (at least initially, and a wonder
must happen so I can chop green from this project), and easier too,
to simply collect WSs problems simply into an Olive file and UL.
A book simply has more...notoriety, is that the word?

1. "Notable" is tricky to define. (Awards and reprints are
at least an objective feature.) What I see as my best problem
hasn't got an award or a reprint. As you said, let the
community decide.

2./3. Seconded, alas, this is work. Unpaid work. Much work.

4. XML? It will take an experienced computer buff ten minutes
to XMLfy PGN. (And probably a year beforehand to discuss
all the features you want :-)
 
   
(Read Only)pid=26196
(6) Posted by Frank Richter [Sunday, Sep 15, 2024 20:31]

4. XML?

Not new ... see https://problem-xml.sourceforge.net/
But don't check "Last modified:" ...
 
   
(Read Only)pid=26198
(7) Posted by Kevin Begley [Monday, Sep 16, 2024 05:44]

@Hauke,

>0. A book simply has more...notoriety, is that the word?

I'd argue the notoriety (good word) of physical books is fleeting.
Digital notoriety will outlast paper notoriety (if the power stays on, no Coronal Mass Ejections, no EMPs, in peacetime, ...).
Err, OK, maybe all notable problems should appear on paper.

>1.
Full agreement here.

>2./3. Seconded, alas, this is work. Unpaid work. Much work.
I have this part largely done (for composers, solvers, MANY OTB/Corr chess players, computer chess players, and even a few persons tangentially related to chess).
I also have genders, a human/computer field, developer links (for people who develop computer chess software/hardware), book links, problem themes (in various languages), variant boards (Hcyl, VCyl, Torus, various holes, variant sizes, etc), fairy pieces, fairy piece movements, fairy piece families, unit colors (w/b/n), time controls, countries (w/ dates existing), flags, fairy conditions, variant games, fairy/variant definitions (in various languages), fairy/variant game inventor linkage, chess quotations, and more.
Plus, I have rating fields for OTB and Corr going back to the first ratings (for top players anyway), date of achieving titles (so you can search youngest/oldest achievement of FM, IM, GM, whatever title you like).
No ratings yet for solvers, but someday I will look for old rating lists for solvers.

Text bios could use improving, but I have some of that right.
I have relative links (ma, pa, bro, sis, cousin, wife, husband, ex-wife, whatever you like).
I have DOB, DOD, birthplace, FIDE ID (for many), ALL possible titles (incl all WC titles, and Olympiad medals) are implemented (but not all titled players added to the DB yet).
Alias names, alternative name spellings, names in other alphabets, etc.
I have a database of problem journals (year started/ended, country, alternative names), and problem composing tournaments (ditto).

I have a very small game db (with events, city, dates, rating category, major tourney or not, etc -- so you can search by Corus, or Olympiads, or list all major tournaments a player has won), and a problem db (fairies and retros take real work, also variant boards are tons of work).
Lots more...

Happy to freely share what I have, if the community decides on a problem format.

Anything I can think might be interesting to search in chess, I try to allow for those fields in a database.
I'm sure there's still a few hundred fields I haven't even considered (and there's many I have forgotten to mention)...

I started working on this for my own sake many years ago, and I have been slowly updating it.
The photos vary (and I can't share those, cuz I foolishly neglected to collect photographer names, and I foolishly forgot to collect only sharable photos in my personal collection).

But, it's probably better to decide on a format for each db (composer db, problem db, etc), and ask problem DBs to dump into the right place (so we can take the best fields from each).

Best solution would be to make it like Facebook (but instead of posting stuff, you just add problems, and update your own personal db).
Why pay law enforcement build a person database, when you can let the people build it for you on a Facebook platform (and sell ads to pay for improvements/data entry)?
If everyone adds their own info, their own problems, their own titles, much of it would build itself.


>4. XML?

I suppose so. I'll let somebody smarter figure that.

An interesting issue is problem storage. Personally, I think you want to store every intermediate fen (in the course of the solution -- to find anticipations, make PG trees, etc).


> [It will take] ... probably a year beforehand to discuss all the features you want...

A year?
No good thing in problem chess happens in less than a decade.
I'm guessing thirty years to discuss and reach an agreement who is authorized to decide these matters. :)

What you really need first is to devise a committee which standardizes chess problems with a strong, unambiguous definition-based classification system (what's a fairy condition, what's a fairy piece, what's a chess variant, what's a stipulation, what are all the possible elements of a stipulation, etc).

When you get all that sorted, it becomes MUCH easier.

But, you can't get anywhere unless you have groups responsible for each database section (not only data input, and determination, but also designing w/ an eye to future innovations).
And, you need some big picture people, with a very sharp eye for future innovations!

Most of all, you need a channel of authority, so you have room to argue about the best way to do this, but not argue forever (which is the best way to stay lazy, and avoid doing anything of value).
You need some group that can make final decisions that everyone will respect, even if it's not what we may have preferred.

And you want to start by building from the least controversial elements, so you have time to adjust to the shape it's taking.

Just think about the FEN you need to universally describe fairy chess problems.
You need a FEN that contains board size information, hole information, it may need to describe all captures awaiting rebirth (e.g., volcanic circe), etc.

You may need somebody to resolve: we're not covering problems bigger than an arbitrary board size (say 26x26, or whatever).
Beyond that, maybe it's just an image (no movement).

There will be tradeoffs: how much can we cover versus what is the impact on database size?
What is a fairy board, versus what is a fairy condition?
Is a hole a unit, or is it a board definition?

A group that understands these issues deeply is needed to make decisions, and not everybody will agree.
So, you need a group that has a strong handle on these matters (people who know how to logical classify problems by definitions, avoiding overlaps/ambiguities).
You need people who understand the long-term implications of these decisions (especially for problem elements not yet invented), so you needn't redo everything!
Most of all, you need people whose decisions can be trusted, to minimize any necessary restructuring.

I'm talking about a very long-term project, which requires at least a decade to design, and a few decades to fill in.
One database format that covers everything in chess, where everyone can contribute to it, and anyone can code their own implementation (based upon one shared standard).

Then, you make some chess social media application, where people can start off by updating their own contributions (and eventually, fill it all in, with help from AI LOL).

And then you get an AI to guide newcomers through it all. :)
 
   
(Read Only)pid=26201
(8) Posted by Hauke Reddmann [Monday, Sep 16, 2024 08:22]

Orrrrr: After 20 years, when the daunting project is finally greenlighted...
the AI is so far developed that it does it in a minute :-)
 
   
(Read Only)pid=26202
(9) Posted by Kevin Begley [Friday, Sep 20, 2024 01:49]

If there was a Chess Problem Composition coin, our motto on it would read: AI or bust. :)
 
 
(Read Only)pid=26214

No more posts


MatPlus.Net Forum Promenade "Problems by ..." - worthsome project?