Latest Version

[Bogg]

Mike:

Its been two months. When are you planning on supplying that bug fix release you promised for September 2016?

Other bugs to report, the epd parameters won't accept realistic values. They were too low in previous versions but at least of limited use. Now they are completely useless. Perhaps 0 for nodes tells it to ignore nodes, if not then a limit of less than 10,000,000,000 is a joke (my two year old computer evaluates 10 million N/S.) A limit on seconds less than 600 also seems ridiculous. I don't know why anyone would search to a ply limit so I guess that should allow something on the order of 2^30 if zero doesn't mean ignore. You can expect the values listed above to be obsolete in the near future so unless this is intentionally planned obsolescence I can't for the life of me think of any reason for your putting limits on any numeric setting anywhere in the program less than 2^63. If a user sets a parameter to some ridiculous value they will learn to read the manual/help, assuming, unlike the current release, it is published.

I'm still looking for an actual improvement in the new version as opposed to old features having been crippled. Saving an occasional keystoke doesn't count.

A suggeted enhancement is to allow backsolving numeric assessments to ignore null values just like informant rate codes.

[MikeAtBookup]

I'm still working on the update. There are four users with Windows 10 who are experiencing a bug that prevents them from clicking on any candidate beyond the first one. Naturally I can't duplicate it. I really want that one fixed even though it is affecting a tiny number of users.

I will check on the EPD analysis limits. The default values are an average setting, way beyond "blunder checking" and probably far short of what most people would want in overnight analysis.

[cfan]

Mike

I appreciate your efforts to fix bugs.

Regarding analysis I agree with Bogg. Users should be able to set whatever limits they please, reasonable or not, within the limits of the computer. Bogg's idea of number = 0 means ignore this limit is a good one. Of course, all this should be made clear in the manual, when we get one!

Some questions and suggestions:

When will speed learning be included?
When will there be a new help file?
Since the Windows 10 issue only affects a few users, why not release a new version while you sort this bug out?
Can you include an option to delete all the comments from the book? This is useful if you want to reanalyse the book positions and delete all previous engine analysis. This has the downside of deleting normal comments, like "from the game A-B, 2016...". A solution to this potential downside is to have two comment areas, one for general comments, and one just for engine output with options to delete all of the comments in either of these areas.

Thank you for considering these ideas.

[Bogg]

cfan

I think you could export the ebook as a game (without comments)

Mike, wasn't there an option to export a database as many games, one for each variation?
if that function is available it would be a better choice.

and import the pgn file into a new ebook.

If you also use CB you could remove the comments there by
1) export the ebook as a game (with comments) See note above
2) open the pgn file as a CB database
3) copy it to a CBH database
4) uncomment the databae.
5) copy the uncommented db to a PGN database
6) import the pgn db into a cow ebook.

If the first idea works it seems to be the simplest and cheapest choice.

Another method that would provide much more flexibility at no additional cost, if you have the requisite skill set is to
1) export the ebook as a game. See note above
2) edit the file with one of the free versions of vi for windows.
3) write a sed command to remove the comments that you want to remove.
4) import the edited file into a new ebook.

If you are unfamiliar with vi let me know and I will see if I can document the exact steps to accomplish your goal (removing the comments from all nonterminal nodes?)

I hope some of that is helpful.

MIKE

Another suggestion is to add a flag to the analyze epd to continue the analysis until the current ply is finished. Stopping mid ply doesn't provide stable information. It has been a long time since I've used the analyze epd feature so perhaps this is already implemented as the standard action. When performing analysis using higher limits this would greatly improve the results at the cost of more time.

[cfan]

Bogg

Thanks for your suggestions. I have Chessbase so your method will work. I still like the idea of having two comment windows, but it's up to Mike if he likes the idea.

Mike

Some further suggestions and comments:

For epd analysis, change the default "Minimum nodes" to be in millions, not thousands, which is too small. Another option instead of 0 = ignore is to have tick boxes next to the three options which when ticked mean ignore this option.

Please include a section in the manual on how to set uci engine options. I only found out how to do this by reading this thread. Also if I want to use multipv and have the engine give more than two top moves, the lower section of the engine output can't be enlarged to show them clearly. It would be good if the variations in the lower section of the engine output are numbered like Chessbase does.

What do "Database Safety" and "Data Version" mean under "About This Book"? Database safety is currently no. Can it be changed to yes?

Thank you.

[MikeAtBookup]

Currently, Database Safety cannot be enabled - and you probably would never want it to be enabled. The SQLite database system is incredibly slow in that mode.

The regular mode is quite robust at the expense of being somewhat slower than the legacy version. I've not seen an ebook yet that has been damaged by a crash.

[Bogg]

Mike:

It has been another two months and still no bug fix update. I can only assume that you are not working on this. Please return my money as you are not acting in good faith.

CTCampbell

[mochess]

I just updated and installed Stockfish 8 64 bit version. It seems to be working fine. I wish (this is picky) the output was not displayed so close to the edge of the window. It feels like it's cutting something off, it's not. Wow 40 million moves and a depth of 25 within a minute!

[MikeAtBookup]

As of December 22, 2016, the latest Professional version for Windows is build 66.

The new version now handles the number of positions in millions (instead of thousands) for overnight analysis. The updated windows show exactly what the minimums are (nodes, time, seconds) during analysis.

[Bogg]

As I mentioned before 10B n/s making a limit of 100 million nodes instead of 100 thousand nodes no improvement and hence still broken.
The ridiculous candidate font size hasn't been fixed. Please remember to place the backsolve info side by side when this is fixed.
Fixing the piece siding speed bug is pleasant but its not like you fixed anything usefull.

In other words you still haven't fixed the things that you promised to have fixed 4 months ago. I'm still waiting for my money back. Make the fixes that were promised 4 moths ago by 2016/12/31 or return my money. My next step is BBB and small claims court.