|Page: [Previous] [Next] 1 2 |
|(1) Posted by Dmitri Turevski [Friday, Oct 10, 2008 09:49]|
Yet another chess problem database
I wrote a few bits of software related to the chess composition, it is all very experimental, but perhaps you could find it useful.
Here's the link:
Feedback is welcome.
|(2) Posted by Harry Fougiaxis [Friday, Oct 10, 2008 10:49]|
Nice and clean interface, pretty fast, most promising. I checked quickly a couple of helpmates : unfortunately neither twins nor solutions are included.
|(3) Posted by Mihail Croitor [Saturday, Oct 11, 2008 00:24]|
solutions exists, just click at problem's id
|(4) Posted by Harry Fougiaxis [Saturday, Oct 11, 2008 07:42]|
...of course, but for some of the entries only.
|(5) Posted by Dmitri Turevski [Saturday, Oct 11, 2008 07:56]|
Yes, indeed, about 2/3 of the problems have solutions (all twomovers and longmovers from Valery Smirnov file and some s#). However, i think i'll run the Popeye on all the problems with reasonable ply count (say 10 halfmoves) and include these solutions too.
There's a trouble with twins, true. Whoever collected the problems (i don't yet know where the file with non-direct problems comes from) they didn't include twins, except for some rare cases when twins were recorded as separate problems (e.g. http://dt.dewia.com/yacpdb/?id=62085 and http://dt.dewia.com/yacpdb/?id=62086). More precisely, i think twins might got lost when someone compiled several smaller collections into the PGN file.
|(6) Posted by Mihail Croitor [Sunday, Oct 12, 2008 19:14]|
i tried to find my problems in this DB. was finded a problem - http://dt.dewia.com/yacpdb/?id=4771
very interesting data about this position (let see "Original PGN"):
[Site "Mat Plus, 2007"]
|(7) Posted by Dmitri Turevski [Wednesday, Oct 15, 2008 14:36]|
I solved (almost) all the problems in 10 or less halmoves with the Popeye. Now 96% of the entries have solutions and there's a new database dump for those who prefer to use the offline version.
This is how the Problemist (the program, not the magazine :) ) exports to PGN, looks like it has no separate field for date. It also discards the twins.
This raises an interesting question: the sources i have used seem to be heavily based on problemist pbm collections, and twins very likely got lost during the PBM to PGN conversion. This can be quite a job reimporting (and solving again) entries with twins from the original pbm files, so i need an advice:
1. Do you think this is worth it (yacpdb with twins)?
2. Do the collections at http://pagesperso-orange.fr/problemiste/collecto.htm really have comprehensive twin information?
|(8) Posted by Dmitri Turevski [Wednesday, Nov 5, 2008 12:38]|
There are some updates to the database. Few thousand new problems, few thousand problems with twins, few fairies (maximummer), solutions and a tool to convert Problemiste files to the Popeye input format (twins sensitive). Everything is free to use and download of course.
|(9) Posted by Mihail Croitor [Wednesday, Nov 5, 2008 20:20]|
Thank You for the great work!
i have a question: i tried to use tch.py and optained an error:
C:Downloadstch-0.1.0>echo "8/4r3/6n1/6Rr/Q2P1k2/p5R1/5N2/Kb6 w - - 0 1" | tch.
Traceback (most recent call last):
File "C:Downloadstch-0.1.0tch.py", line 2858, in <module>
File "C:Downloadstch-0.1.0tch.py", line 2846, in main
for line in input:
IOError: [Errno 9] Bad file descriptor
What i maked incorrect?
|(10) Posted by Dmitri Turevski [Wednesday, Nov 5, 2008 22:08]|
This is a well known bug in windows cmd.exe, it can't handle the piping properly. You should have followed the tch INSTALL doc when working with windows and give it the full path to the python executable, e.g.:
C:\Downloads\tch-0.1.0> echo "8/4r3/6n1/6Rr/Q2P1k2/p5R1/5N2/Kb6 w - - 0 1" | C:\Python2.5\python.exe tch.py
I don't have a windows machine at hand right now, but i remember having the same issues.
|(11) Posted by Mihail Croitor [Thursday, Nov 6, 2008 02:37]|
thank you, it work now
|(12) Posted by Hauke Reddmann [Thursday, Nov 6, 2008 16:30]|
Doing a few searches, I immediately spotted a musthave for
a task freak like me: upgrading "Exists[$Theme]" to "Numberof[$Theme]" :-)
|(13) Posted by Dejan Glisić [Monday, Dec 15, 2008 19:28]|
Dear Dimitri, do you need some problems for your database? I can send you some collections from Croatian sources and authors.
|(14) Posted by Miodrag Mladenović [Monday, Dec 15, 2008 21:44]|
I look today for the first time this new database and of course, first I queried for my problems just to get an idea how many problems query will return. I've been pleasantly surprised when it returned all of my twomovers. I cannot find any problem missing.
However there is a problem listed as a joint composition (by me and Mr. Klemanic):
This is not my problem and my name should be removed from the list of authors. I cannot take any credit for this problem. I did compose different one with slightly different matrix that won 1st Prize at Die Schwalbe.
Anyway, excellent work. The results came on the screen very quickly.
|(15) Posted by Dmitri Turevski [Tuesday, Dec 16, 2008 10:14]|
Hello Dejan, yes, of course i would be glad to add your collection. You didn't really have to ask, all contributions/corrections/ideas are welcome. Please send me your files.
Miodrag, thank you for the appreciation of my work. I have fixed the entry you have referenced. You may find it interesting that i have also looked up this problem in another database (the one with happy users) and it also gives you and Mr. Klemanić credit for this problem as well.
Hauke, it is a wonderful idea of yours, but unfortunately, after giving it some thought, i find it a little beyond my skill. The problem is that in many cases the same effect/theme can be doubled (tripled etc) in more than one way. So, first, some extendible, flexible and comprehensive classification system must be produced. And at the moment i believe that having no such system could be better than having a poor one that i can come up with.
|(16) Posted by Dejan Glisić [Tuesday, Dec 16, 2008 11:51]|
Sorry, I forgot your e-mail address. Please contact me to: pzrdig2-at-gmail.com. Best regards!
|(17) Posted by Dmitri Turevski [Wednesday, Dec 31, 2008 18:21]|
The database has been updated. There are 23000+ more problems to browse now and a new dump for offline users, as usual.
Happy New Year!
|(18) Posted by Dmitri Turevski [Thursday, Mar 12, 2009 10:37]|
I recently wrote a simple wiki engine for yacpdb. This means that now everyone (you!) can edit every problem in the database, fix diagrams and solutions, correct sources and authorship, leave comments on problems, add trivia, include bits of info on your problems that were not published in magazines, discuss judgment, whatever, everything. And you can check everyday for what new was submitted since you have last checked.
Wikis work in such a way that no contribution is ever lost, all older versions can be restored with just few clicks. And you can not do any harm with your edits, only good. Please feel free to "play around", i really tried to make it user friendly, there's even no need to register. Here's a point to start with more info:
I invite everyone to contribute and discuss. The idea was born when Rainer Staudte pointed out that sending inaccuracies to me by email is obviously not a perfect way of correcting the errors.
The success of such a community project depends wholly on contributors, it'll get improved and developed but only if there's interest.
|(19) Posted by Siegfried Hornecker [Thursday, Mar 12, 2009 12:49]; edited by Siegfried Hornecker [09-03-12]|
It seems you also need "" on comments if you use a : in it. Updated my problems with solution and anticipation info.
Is there a way to include new problems? I'd like to include my pawn-rundlauf-fivemover with Gilles Regniers.
PS after reply of DT: So since there isn't, maybe you want to do? -> http://chessproblem.net/viewtopic.php?f=44&t=296#p658
|(20) Posted by Dmitri Turevski [Thursday, Mar 12, 2009 13:32]|
Yes, you are correct, in this case the parser interprets the colon as if you were defining a nested dictionary. However, it may be more convenient to mark comments as block literals with '>' or '|' characters. Check how the yaml-library converted your comment with colon:
Also one of the things on my todo list is to make a separate form for adding a single comment (and more generally - for separate edition of each section), so you don't have to worry about such things.
And submitting new problems, yes, this is on TODO too. And fairy pieces.
|Page: [Previous] [Next] 1 2 |
MatPlus.Net Forum Internet and Computing Yet another chess problem database