Fritz5: bugs & wishes, summary
Heiko A. Groeneveld
Since there appears to be an end of the discussion concerning my bug&wishlist
in sight _and_ because MW is back from his vacation and participating now,
I'd like to summarize what we've got -- from my point of view, of course.
Above all, I'd like to stress two points:
1) Fritz4/5 and CB are important tools for me, mostly because they support
variants, what all other chess programs that I am using regularly do not --
i.e. -- ChessGenius, XBoard/GnuChess/Crafty, WChess, MChess (listed in the
order of time I'm spending with them). So the intention of this list is
entirely constructive: the list is meant for the information of other users
as well as for the information of ChessBase in order to help improving
Fritz/CB.
2) The Fritz5 crashes I reported are tracked down to two reasons:
a) A kind of software/hardware incompability between Fritz5/W95
and my IBM/Cyrix P166+ processor. After replacing the Cyrix
with an AMD, no crash out of the blue happened during a whole
week -- whereas I had at least one per day before. It remains
an open question where exactly the problem is located: Is it
the processor, is it W95, is it Fritz? Experiences of other
Cyrix users welcome.
b) The ScrollBug-crashes I reported (see also below) are reproducable
on my system even with the AMD, but are admittedly much more
difficult to reproduce than it has been with the Cyrix.
In the following list, I am partially citing MWs answers to my original list
and marked my replies to him with periods at the beginning of each line.
Please note, that the wish to have Drag & Drop in the ``set up position''
window vanished from the list, because MW e-mailed me that he already
added this feature. Many thanks (in advance, since I don't have it yet)! :-)
Of course, some new points appear in the list. Thanks to all who contributed!
If you Fritz users out there encourage me, I can make this list a WWW-page.
Bugs:
-----
o The red warning LED is malfunctioning in Fritz5
(Harald Faber was the only person reporting it works correct on his system)
o The ScrollBars/Sliders are buggy:
- It's possible to ``scroll'' an empty window by clicking on the
single-step button below the slider
- By clicking below the slider, Fritz5 isn't scrolling through the
notation, but through the game (the selected move changes)
MW: Yes. This is the way to quickly scan a game.
. That's declaring this bug a feature. I don't know, if that's a good
. policy. When I'm in analysis mode (that's where I am most time when
. using Fritz) and would like to flip through the notation (because
. I am analysing in parallel in my own head) it's _very_ inconvenient
. that I can't scroll the notation pagewise without disturbing Fritz's
. analysis. And it's also inconsistent, like MANY other things in Fritz5.
- By clicking on/below/above slider and on the slidebuttons heavily,
it's possible to confuse and even crash Fritz5
I gave recipes for this in former threads. Everybody was able to
reproduce the confusion, nobody reported so far that he reproduced
a crash.
o The ``reddening'' of the buttons in the main window as well as in the
database window doesn't coincide with their selectability.
(Again, Harald Faber couldn't reproduce this, MW couldn't yet, too.)
Inconveniences / Strange effects:
---------------------------------
o After replacing a game in the database window, another game
(the top one in the list) will become selected
MW: Yes. Needs to be changed.
o The buttons Fritz4/Fritz5 below Brett/2D/Einstellungen
(that's probably board/2D/options in the English menu)
work different.
MW: Yes. This is inconsistent. We will implement a "Select color set" function.
o The new display of the hash filling state shouldn't get red, when the
hash overflows. The red colour irritates the player by making him think,
that the red warning LED flashes. At least, it irritates me.
MW: Should I make it yellow?
. I'd prefer the same camouflage-colour as used for displaying the
. percentage. The word FULL is to my mind striking enough.
o The buttons in the database window are placed badly at the bottom
of the window, because
a) the distance from list entry to button is longer this way
b) the list is shorter
c) it's harder to group the buttons
and I'd love to have them grouped in some sensible way!
MW: On smaller screen resolutions, it is important to have the maximum
screen width available. Please suggest a better grouping for the
buttons. Other users have complained about this as well. Somebody
suggested menus, but in my opinion this means more mouse click hazzles.
. I am using 800x600 and for me the WIN of window width doesn't make up
. for the LOSS of window height. I don't believe that there are many
. 640x480 users out there, except perhaps on notebooks. I'll think about
. a grouping scheme and will try to post my suggestion next week.
By the way:
o The horizontal scrollbar in the database list doesn't represent the
amount of horizontal information in the list. It should disappear
entirely if all horizontal information can be displayed.
o Sometimes, Fritz5 saves the last used engine, sometimes, it doesn't.
MW: If you have loaded more than one engine, only the most recently used
engine is saved. Beginners might otherwise clutter their memory with
loaded engines.
. So make it an option (for us ``non-beginners'').
o When leaving Fritz5, sometimes the actual game is kept and restored when
restarting Fritz5, sometimes not. I'd like to have it kept _always_
MW: Its being stored if it is a game you played. Its not being stored if
it is a game loaded from a database. I understand your point.
. Play a game. Leave Fritz. It stores the game (most times, see below).
. Restart Fritz -- the game is back -- leave and restart once again --
. the game is gone.
. Play a game. Leave Fritz while still in book. Restart Fritz. The game
. is gone. Consistency?
. Well, you already understood my point. ;-)
o The documentation of several features is poor
MW: Please be more specific.
. Really, I'd love to be, but this already took me more than three hours.
. To be more specific here, would cost me a day.
. BTW: Giorgio Bertazzo and others felt the same way.
. I suggest that you give Fritz and the documentaion to a newbie
. and watch him stumbling, that would be probably very instructive.
. Just to mention one point: The new ``correspondance-chess analysis''
. feature needs _exhaustive_ documentation, preferably with _examples_.
o There should be a warning if cbh features aren't stored in a cbf file
o There are doubts that there is 100% compatibility between cbh files
of CB6 and those of Fritz5. More experiences to this point needed
and welcome! (I still didn't find the time to install CB6 myself.)
o In Fritz4, the current book-moves were displayed in the analysis window
while playing a (non-serious) game, in Fritz5 (mine, at least) this
works only for the first move
Wishes:
-------
o An optional HEUMAS
MW: Heumas is a double-edged thing in Fritz. Some users don't want it at
all. You can switch it off by setting "MiniHeumas = 0" in F5.INI. If you
set "MiniHeumas = 2" it will become more aggressive in guessing.
. Why not place a button somewhere to change this interactively with Fritz?
. Does this ``MiniHeumas'' concern the ``serious'' mode, too?
. I'd like an ``intelligent'' Fritz Heumas, rather a ``MaxiHeumas'', ;-)
. where Fritz himself suggests the best move with the clicked piece.
o When running a long analysis, and something more important than chess
has to be done on the same computer, it would be a pleasure to have
a ``Freeze'' button, enabling the user to store the whole state of
Fritz5, including hash. After restarting Fritz, pressing the ``Melt''
button would just continue the analysis where it stopped.
MW: Good idea. Doesn't sound easy.
. A challenge. :-) Keep it for Fritz6. ;-)
o Voice input / output for playing blind(fold) games
MW: Do you know a good voice input development kit? I don't want to
program this myself.
. The reactions prove interest for this feature! Sorry, I don't know
. a voice development kit, but CrativeLabs most probably have something.
. (It can't be too difficult for from-to notation, I saw voice input
. for moves like d2-d4 programmed on the Apple ][ ten years ago.)
o Higher degree of desktop configurability, more consistency
Take, e.g., the new ``view book'' function. Entering and leaving
it resets the size of the notation window. Why?
And finally, from the programmers guild: ;-)
o Documentation of cbf and cbh file formats
o Documentation of the engine interface
Yours in chess,
+ Heiko +
P.S.: I'll be rgcc-offline for approximately a week, but will read my e-mail
--
Harald
quoting a mail from zra...@linus.zdv.uni-tuebingen.de
> Since there appears to be an end of the discussion concerning my
> bug&wishlist in sight _and_ because MW is back from his vacation and
> participating now, I'd like to summarize what we've got -- from my point of
> view, of course.
> Bugs:
> -----
>
> o The red warning LED is malfunctioning in Fritz5
> (Harald Faber was the only person reporting it works correct on his
> system)
The red warning LED is only present if I go into a database window, select
a game (or more) and let Fritz5 analyze it.
But it isn't bad at all because it stays red the whole time the game is
analyzed. And most do it overnight I suppose. I myself don't want to look
at the not understandable running analysis.
> o The ScrollBars/Sliders are buggy:
> - It's possible to ``scroll'' an empty window by clicking on the
> single-step button below the slider
> - By clicking below the slider, Fritz5 isn't scrolling through the
> notation, but through the game (the selected move changes)
>
> MW: Yes. This is the way to quickly scan a game.
But not correct. If I click a little below the slider I want to have a
look a little further, not look at the end.
> o The ``reddening'' of the buttons in the main window as well as in the
> database window doesn't coincide with their selectability.
> (Again, Harald Faber couldn't reproduce this, MW couldn't yet, too.)
Tried again, absolutely correct. If the mouse pointer leaves the button
area the buttons get black.
Maybe we should add the problem with Win95 and the figurine-symbol-
fonts...
> Inconveniences / Strange effects:
> ---------------------------------
> o The new display of the hash filling state shouldn't get red, when the
> hash overflows. The red colour irritates the player by making him think,
> that the red warning LED flashes. At least, it irritates me.
>
> MW: Should I make it yellow?
>
> . I'd prefer the same camouflage-colour as used for displaying the
> . percentage. The word FULL is to my mind striking enough.
I am not affected by that. Although it could be "100%" or "full" which is
more humble.
But this is only cherry-picking.
> o The buttons in the database window are placed badly at the bottom
> of the window, because
> a) the distance from list entry to button is longer this way
> b) the list is shorter
> c) it's harder to group the buttons
> and I'd love to have them grouped in some sensible way!
What about adapting some other idea. It is done I don't know in which
program. You can choose if you want to have the icons on the left side, on
the top, on the right side or at the bottom.
> o The horizontal scrollbar in the database list doesn't represent the
> amount of horizontal information in the list. It should disappear
> entirely if all horizontal information can be displayed.
Agree. Then a vertical button ordering like in Fritz4 would make much more
sense!
> o When leaving Fritz5, sometimes the actual game is kept and restored when
> restarting Fritz5, sometimes not. I'd like to have it kept _always_
>
> MW: Its being stored if it is a game you played. Its not being stored if
> it is a game loaded from a database. I understand your point.
>
> . Play a game. Leave Fritz. It stores the game (most times, see below).
> . Restart Fritz -- the game is back -- leave and restart once again --
> . the game is gone.
> . Play a game. Leave Fritz while still in book. Restart Fritz. The game
> . is gone. Consistency?
> . Well, you already understood my point. ;-)
This is what I absolutely don't understand. All the other programs keep
the last position saved or whatever so that a new start gives the last
position you had on the board, no matter where it came from.
> o The documentation of several features is poor
>
> MW: Please be more specific.
>
> . Really, I'd love to be, but this already took me more than three hours.
> . To be more specific here, would cost me a day.
> . BTW: Giorgio Bertazzo and others felt the same way.
Without being specific. So no possibility to improve.
> o In Fritz4, the current book-moves were displayed in the analysis window
> while playing a (non-serious) game, in Fritz5 (mine, at least) this
> works only for the first move
Right-click in the "analyze"-window and select "Buch immer pruefen". It is
kept even if you leave F5 and start again. :-)
And shows the actual book moves.
> Wishes:
> -------
> o When running a long analysis, and something more important than chess
> has to be done on the same computer, it would be a pleasure to have
> a ``Freeze'' button, enabling the user to store the whole state of
> Fritz5, including hash. After restarting Fritz, pressing the ``Melt''
> button would just continue the analysis where it stopped.
>
> MW: Good idea. Doesn't sound easy.
Will probably cause memory/hash problems.
> o Voice input / output for playing blind(fold) games
Seems to me you want an "eierlegende Wollmilchsau" (however I may
translate it into english, I don't think there is a word for that) :-)
As I told Matthias the EPD-import, -export and -analysis function is
missing. He promised to add those.
I also wish, for there is a database function implemented, a non-
restricted database-function (no limit in search to 4999 games!)
Harald Faber