scidb-users Mailing List for Scidb

Pietro wrote:
> No problem for the database, but I think to have isolated the problem, so it
> may be interesting for you: my guess regarding the crash is that it happen
> because I wrote the name of the black player with some illegal pattern. The
> name in point was "8.f4, 8.0-0".
> Later I tried without the blank space and it worked fine.

Ok, this was a misunderstanding, the database is not lost.

Unfortunately I cannot reproduce this problem with the pattern
"8.f4, 8.0-0". It's only a string, and the application normally don't
cares about the content of the string.

By the way: I will try to solve this problem with the crash. It's in fact a
result of stack back-tracing after an exception is thrown, normally the
application should show an error message instead. With older compiler
versions this works fine, but with newer versions something has changed
with the stack format, and the back-tracing is not working properly
anymore, sometimes (but not every time) it is crashing. This is very
annoying, because stack back-tracing is a basic debugging technique.
And I cannot find any information in the Internet about this problem.
Anybody else who knows something about this?

Hi Pietro,

>  Do you mean via wine? If not, how did you do?

Yes, with Wine. It works ok, except the thing with the activation.

> I was just using scidb when it crashed and the database i was writing
> went lost. I did not yet restart the pc, is there any chance to recover it?

At first, I don't know a way how to recover lost files.

I'm wondering that the database is lost. The game data file will never be
rewritten - except compaction, and even compaction is writing a copy.
The same applies to the index file, either a new index record will be
appended, or one record will be replaced. Only the namebase file will be
rewritten, but in this case Scidb is writing at first a temporary file, and
after success this file will be renamed to the original file; this file renaming
is an atomic operation and can never fail. Can you describe exactly what you
did?

Pietro wrote:
> No problem for the database, but I think to have isolated the problem, so it
> may be interesting for you: my guess regarding the crash is that it happen
> because I wrote the name of the black player with some illegal pattern. The
> name in point was "8.f4, 8.0-0".
> Later I tried without the blank space and it worked fine.

Ok, this was a misunderstanding, the database is not lost.

Unfortunately I cannot reproduce this problem with the pattern
"8.f4, 8.0-0". It's only a string, and the application normally don't
cares about the content of the string.

By the way: I will try to solve this problem with the crash. It's in fact a
result of stack back-tracing after an exception is thrown, normally the
application should show an error message instead. With older compiler
versions this works fine, but with newer versions something has changed
with the stack format, and the back-tracing is not working properly
anymore, sometimes (but not every time) it is crashing. This is very
annoying, because stack back-tracing is a basic debugging technique.
And I cannot find any information in the Internet about this problem.
Anybody else who knows something about this?

Hi Pietro,

>  Do you mean via wine? If not, how did you do?

Yes, with Wine. It works ok, except the thing with the activation.

> I was just using scidb when it crashed and the database i was writing
> went lost. I did not yet restart the pc, is there any chance to recover it?

At first, I don't know a way how to recover lost files.

I'm wondering that the database is lost. The game data file will never be
rewritten - except compaction, and even compaction is writing a copy.
The same applies to the index file, either a new index record will be
appended, or one record will be replaced. Only the namebase file will be
rewritten, but in this case Scidb is writing at first a temporary file, and
after success this file will be renamed to the original file; this file renaming
is an atomic operation and can never fail. Can you describe exactly what you
did?

 Do you mean via wine? If not, how did you do?

P.S. I was just using scidb when it crashed and the database i was writing went lost. I did not yet restart the pc, is there any chance to recover it?

  
    

I finally managed to run ChessBase 11 under Linux, but the
activation fails. Does anybody know how to activate CB under
Linux? It seems that the detection of the hardware is a problem
for CB (it is sending the hardware information to any server while
activating).

Cheers,
Gregor

PS: I've reduced the images sizes of the themes, used by the
theme voting at http://scidb.sf.net/themes/, about 50%, so the
load of the pages is quite faster now.

PPS: Currently Sourceforge is down (18.02, 13:24).

Drew replied:
> A60 is 1.d4 Nf6 2.c4 c5 3.d5 e6 [4.Nf3] which is a transposition of
> 1.d4 Nf6 2.c4 e6 3.Nf3 c5 4.d5

But this is not really answering my question. If "1.d4 Nf6 2.c4 c5 3.d5 e6
4.Nf3" is A60, then logically "1.d4 Nf6 2.c4 c5 3.d5 e6 4.Nf3 b5" should also
be A60, but it is E10. Note that a delayed b5 will also be assigned to A60,
e.g. "1.d4 Nf6 2.c4 c5 3.d5 e6 4.Nf3 g6 4.g3 b5". With other words the (equal)
lines

  1.d4 Nf6 2.c4 e6 3.Nf3 c5 4.d5 b5
  1.d4 Nf6 2.c4 c5 3.d5 e6 4.Nf3 b5

are an exception because they are belonging to E10, but any other
line starting with either "1.d4 Nf6 2.c4 e6 3.Nf3 c5 4.d5" or
"1.d4 Nf6 2.c4 c5 3.d5 e6 4.Nf3" , not immediately followed by b5, will
be assigned to A60. Why this exception?

On Fri, 13 Feb 2015 17:05:57 +0100
Gregor Cramer <re...@gm...> wrote:

> I've detected a curiosity in the ECO file of Scid:
> 
> 	1.d4 Nf6 2.c4 e6 3.Nf3 c5				will be
> assigned to E10 1.d4 Nf6 2.c4 e6 3.Nf3 c5 4.d5		will be
> assigned to A60 1.d4 Nf6 2.c4 e6 3.Nf3 c5 4.d5 b5	will be

A60 is 1.d4 Nf6 2.c4 c5 3.d5 e6 [4.Nf3] which is a transposition of 
1.d4 Nf6 2.c4 e6 3.Nf3 c5 4.d5

> assigned to E10
> 
> This is not looking logical. Does anybody have an explanation for
> this? A line like "1.d4 Nf6 2.c4 e6 3.Nf3 c5 4.d5 a6" will be assigned
> to A60, and this is logical. Why this exception for 4...b5?
> 
> Cheers,
> Gregor
> 
> 
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming. The Go Parallel Website,
> sponsored by Intel and developed in partnership with Slashdot Media, is
> your hub for all things parallel software development, from weekly
> thought leadership blogs to news, videos, case studies, tutorials and
> more. Take a look and join the conversation now.
> http://goparallel.sourceforge.net/
> _______________________________________________ Scidb-users mailing list
> Sci...@li...
> https://lists.sourceforge.net/lists/listinfo/scidb-users



-- 
Drew Ferguson
AFC Commercial
http://www.afccommercial.co.uk

I've detected a curiosity in the ECO file of Scid:

	1.d4 Nf6 2.c4 e6 3.Nf3 c5				will be assigned to E10
	1.d4 Nf6 2.c4 e6 3.Nf3 c5 4.d5		will be assigned to A60
	1.d4 Nf6 2.c4 e6 3.Nf3 c5 4.d5 b5	will be assigned to E10

This is not looking logical. Does anybody have an explanation for
this? A line like "1.d4 Nf6 2.c4 e6 3.Nf3 c5 4.d5 a6" will be assigned
to A60, and this is logical. Why this exception for 4...b5?

Cheers,
Gregor

I've checked in a small update (r1022), this revision contains all the themes
and piece sets presented in the voting tour on Scidb's homepage:
<http://scidb.sourceforge.net/themes.html>.

This is from the Changelog:

- Fixes and improvements in piece renderer.

- Enhancement in square rendering.

- New piece sets: Aquarium, Bookup, Chess7, ChessCube, Celtic, Fritz,
   Habsburg, Military, Old Style, Segoe, and Zurich.

- Piece set "Standard" replaced by another version, the previous one has
   deficiencies.

- New themes: Burnt, Fritz, Kitsch, Navajo, Marmorate, Melamine, and
   Military.

- Theme "Virtual " re-designed and renamed to "Virtual - Brown".

- Theme "Virtual - Colored" renamed to "Virtual - Blue".

- Themes "Marble - Brown" and "Marble - Red": piece set Merida replaced by
   Habsburg.

- Theme Blackjack: piece set Usual replaced by Segoe.

- Some piece sets fine adjusted.

Cheers,
Gregor

Scidb's homepage has some new pages, any user now may
give a vote to Scidb's board themes.

See <http://scidb.sourceforge.net/themes.html> if you
like to participate.

It was a lot of Javascript, especially JQuery, and some PHP
programming. Please report about errors, it would be helpful
if you can remember the last action before the error has
occurred.

Cheers,
Gregor