Banksia GUI released

[phhnguyen]

To graphic designers and anyone who can help:

I wish to have a new logo for Banksia GUI.
The requirements:
- square, size from 1024x1024 (or even larger)
- has a large light color area (background or content area) thus it can be seen easily when it is in menus, bars, desktop screens of multi OSs (Windows, macOS, Linux)
- simple, clear to see, easy to recognize and distinguish (with other logos)
- no text or very few characters

I have a small budget of $US100 as a little pay/bonus to someone whose logo I can use (fully or partly).

You may post here or send it to me via email.

Any help will be highly appreciated!

[Damir]

Ovyron is one of the best logo makers out there.. I am sure he will be able to help you.

[Jonathan003]

The time for creating a PGN opening book depends on computers, the stronger one will be faster. On my 2-year-old computer (3.6 GHz Quad-Core Intel Core i7), it took about 5 minutes to convert a game database of 1.7 million games (human GMs) into a PGN opening book. It means 800 K games will take roughly 2-3 minutes.

Image

Thanks for the explanation, but my question was how long it will take to convert the big pgn file to bdgb format. I already have tried it and it took about 2 minutes to convert the complete database. I set the options for min game length to 1, min score to 0, and set the tread number to 4. (I have 4 cores on my machine so I think to let it use 4 threads is a good choice).
I get this message at the bottom after conversion to bdgb format: games: 781885, added: 135, period: 01:50, spd: 7108
What does this mean? how comes only 135 games are added? And what does spd means? I find it suspect that the pgn file with the 781885 games is 554,224 KB big. And after conversion the bdgb database is only 31 KB big.
I also have a question about the book making option. What does the Random+- settings mean? And what are the best settings for this?
I created a pgn book from the converted database. I used these settings: min player score 0, min game length 1, repeat at least 1. Add to: 99 plies and random +- 2
I got this message: games: 135, period: 00:00, spd: 135
If I open this pgn book in a text editor. The notation doesn't make any sense. It looks like the moves are not always starting from the starting position. I want to use these pgn books for repertoire making so this is useless for me.

Here is a small sample:
[Event "1"]
[Site "?"]
[Date "????.??.??"]
[White "?"]
[Black "?"]
[Result "*"]

1. Bh3 f6 2. Re3 Bg8 3. Nb6 Rd1+ 4. Kf2 Rd2+ 5. Re2 Rxe2+ 6. Kxe2 Nd4+ 7. Ke3 Nc2+ 8. Kd2 Nxa3 9. Nxa4 Nc4+ 10. Kc3 Ne5 11. Bg2 b5 12. Nc5 Ke7 13. f4 Nc6 14. Nd3 Kd6 15. Bf3 Bc4 16. Nb4 Nxb4 17. Kxb4 g5 18. h4

[Event "2"]
[Site "?"]
[Date "????.??.??"]
[White "?"]
[Black "?"]
[Result "*"]

1. Bxe6 Nxe6 2. Nf5 Kh7 3. Qg3 Bf8 4. Qh4 Ng8 5. Ne2 g6 6. Ne3 Be7 7. Qg3 Nf6 8. Nc4 Nxe4 9. Qe3 N6g5 10. Nc3 Bc5 11. Nxe4 Bxe3 12. Nf6+ Kg7 13. Nxe8+ Rxe8 14. Bxe3 b6 15. Rad1 f6 16. Rd3 Rd8 17. Red1 Rxd3 18. Rxd3 Nf7 19. b3 h5 20. f3 Qe7 21. Bf2 Qc7 22. Rd1 Qe7 23. a5 bxa5 24. Nxa5 Qc7 25. Nc4 a6 26. Bc5 Qb7 27. Ra1 Qd7 28. Kf2 Qf5 29. Ne3 Qd7 30. Rd1 Qb7 31. Nc4 Qb5 32. Be3 Qb7 33. Nb6 Qe7 34. Rd7 Qb4 35. Nc4 Kg8 36. Ra7 Qc3 37. Na3 Qa5 38. Nc4 Qc3 39. Na3 Qa5 40. Nc4 Qc3 41. Ra8+ Kh7 42. Rxa6 Qxc2+ 43. Nd2 Kg7 44. Ke2 f5 45. Ra7 h4 46. b4 Kf6 47. Bc5 Kg7 48. Rc7 Qa4 49. Rb7 Qa6+

[Event "3"]
[Site "?"]
[Date "????.??.??"]
[White "?"]
[Black "?"]
[Result "*"]

1. Kb4 Kd6 2. Rb1 Rf3 3. Rb3 Rxg3 4. c5+ Kc7 5. Kc4 Rxh3 6. Rb6 g5 7. Re6 g4 8. Rxe5 g3 9. Re7+ Kc8 10. Rg7 h5 11. a5 h4 12. a6 bxa6 13. Kd5 Rh1 14. Kd6 Rf1 15. Rg8+ Kb7 16. Rg7+

[Event "4"]
[Site "?"]
[Date "????.??.??"]
[White "?"]
[Black "?"]
[Result "*"]

1. Kh2 Kg6 2. Kg3 Kf6 3. h4 Rc5 4. Ra7 h5 5. Kf3 g6 6. g3 Kf5 7. Rf7+ Ke5 8. Rg7 Kf6 9. Ra7 g5 10. hxg5+ Rxg5 11. Ra8 Rc5 12. Kg2 Ke5 13. Kh3 Ke4 14. Kh4 Kf3 15. Rf8+ Kg2 16. g4 hxg4 17. Kxg4 Rc3 18. Ra8 Rc4+ 19. Kg5 e5 20. Re8 Rxa4 21. Rxe5 Kf3 22. Kf5 Rh4 23. Re1 a4 24. Ke5 a3 25. Kd5 Ra4 26. Kc5 a2 27. Ra1 Ke3 28. Kb5 Ra8 29. Kc4 Kd2 30. Kb3 Rb8+ 31. Kc4 Rb2 32. Rh1

[phhnguyen]

Thanks for the explanation, but my question was how long it will take to convert the big pgn file to bdgb format. I already have tried it and it took about 2 minutes to convert the complete database. I set the options for min game length to 1, min score to 0, and set the tread number to 4. (I have 4 cores on my machine so I think to let it use 4 threads is a good choice).
I get this message at the bottom after conversion to bdgb format: games: 781885, added: 135, period: 01:50, spd: 7108
What does this mean? how comes only 135 games are added? And what does spd means? I find it suspect that the pgn file with the 781885 games is 554,224 KB big. And after conversion the bdgb database is only 31 KB big.

The text means that BSG has processed 781885 games but found and added only 135 games and processing speed (spd) is 7108 games per second. A very large number (781885 - 135 = 781750) of games is totally discarded.

Look like games in your PGN file are in different chess variants and/or seriously wrong format since BSG can tolerate with some wrong on PGN files.

You may check by following steps:
- Open your PGN file by any text editor
- Copy randomly a game then paste into BSG

If BSG says it can't open, check what is wrong with that game. If you see nothing wrong please post the game here or send me via email to verify.

I also have a question about the book making option. What does the Random+- settings mean? And what are the best settings for this?
I created a pgn book from the converted database. I used these settings: min player score 0, min game length 1, repeat at least 1. Add to: 99 plies and random +- 2
I got this message: games: 135, period: 00:00, spd: 135
If I open this pgn book in a text editor. The notation doesn't make any sense.

random +- 2 means random +/- 2 moves for the length to take from a game. It is useful if you want to add some random factor to the openings, says, if you get bored when every opening lines have the same length and started by a fixed side (caused by the fixed length).

For opening books only moves are master and PGN tag-contents are totally useless.

It looks like the moves are not always starting from the starting position. I want to use these pgn books for repertoire making so this is useless for me.
...

Thanks a lot for the report. I am going to check and fix the bug soon.

[Jonathan003]

Look like games in your PGN file are in different chess variants and/or seriously wrong format since BSG can tolerate with some wrong on PGN files.

You may check by following steps:
- Open your PGN file by any text editor
- Copy randomly a game then paste into BSG

If BSG says it can't open, check what is wrong with that game. If you see nothing wrong please post the game here or send me via email to verify.

I don't think there is something wrong with the pgn database. Its from high quality databases from Chessbase. I also checked the games for non strandart startpostions. And I cleaned the pgn with SCID

I will send you a PM with a link to the pgn database on my google drive.

[Ovyron]

Sure thing! The hardest part about logo design is having a good idea, if the idea is bad then no good execution can save it.

Here's my attempt at a Banksia GUI Logo:

[phhnguyen]

Look like games in your PGN file are in different chess variants and/or seriously wrong format since BSG can tolerate with some wrong on PGN files.

You may check by following steps:
- Open your PGN file by any text editor
- Copy randomly a game then paste into BSG

If BSG says it can't open, check what is wrong with that game. If you see nothing wrong please post the game here or send me via email to verify.

I don't think there is something wrong with the pgn database. Its from high quality databases from Chessbase. I also checked the games for non strandart startpostions. And I cleaned the pgn with SCID

I will send you a PM with a link to the pgn database on my google drive.

Thanks a lot. I got it. The bug is fixed for the next release (it turns out the regular expression which I have used for my PGN parser will ignore the games with empty tags, such as [Annotator ""]).

[phhnguyen]

Sure thing! The hardest part about logo design is having a good idea, if the idea is bad then no good execution can save it.

Here's my attempt at a Banksia GUI Logo:

Thanks a lot for the logo. It looks so nice!

I agreed, the hardest and importest is about ideas!

I prefer a simpler logo that has only few details and/or shapes.

[Jonathan003]

Thanks a lot. I got it. The bug is fixed for the next release (it turns out the regular expression which I have used for my PGN parser will ignore the games with empty tags, such as [Annotator ""]).

Thanks I'm looking forwards to the next release of Banksia. Is it already possible to use the book merging options for books in pgn format? If not I would like to request this.

[Jonathan003]

Hi, Mr Nguyen Hong Pham

I have some requests to implement in future versions of Banksia.
Chess Position Trainer has some unique way to navigate tro a repertoire.
It is possible to jump to positions where there is more than one candidate move for the side of the repertoire. (> candidate).
It is a unique feature for cpt at the moment.
The problem with cpt is that if I jump to a next position with more than one candidate moves, and I delete some candidate moves. CPT often jumps again to the first position with more than one candidate move, when hitting next.
For example, let's say there are 200 positions with more than one candidate moves, for the side of the repertoire, (you can see the nr of position with more than one candidate moves at the bottom in cpt). And I'm at position 99, and delete two of the three candidate moves. And then I hit jump to " > candidate", there is a big change that cpt will jump back to the the first position with more than one candidate move. And I have to hit 99 times on " > candidate" to continue, from where I was. That's very frustrating.
(It would be interesting if cpt could calculate on the fly how many positions there are left with more than on candidate move, after deleting variations, without having to jump to the first position).
I tried to solve this by creating two openings in cpt, one to navigate, and one to keep. I click on "> candidate" in "to navigate" and I delete the variations I don't want to keep in "to keep"
But that also gives problems in cpt. The variations that are deleted still are shown. I have especially problems with bigger repertoire files in cpt, (it becomes very unstable with big repertoire files). I want to start with a big repertoire and then delete candidate moves I don't want to play or I don't expect my opponent will play often.

I have made a video capture of it so you can better understand what I mean:



CPT also has an option to jump to leaf nodes and ad a engine evaluation, (manually or with an engine). Than there is also an option to jump to leaf nodes without a engine evaluation. I would like it if you could implement this in Banksia. Also to be able when in a certain position in opening lines to copy the moves leading to the position, (to check them for opening errors manually).
Here is a screenshot of a part of the manual of cpt that explain different ways of navigating tro a repertoire.



CPT 6 was promised to come out in 2019 according the facebook page of cpt. But it's still not here.

You can download the manual of cpt here: http://www.chesspositiontrainer.com/dow ... Manual.pdf

GRT