[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-bugs-dist
Subject:    [Bug 279229] Solver could be made a lot more efficient with 'undo' ...
From:       Parker Coates <parker.coates () kdemail ! net>
Date:       2011-11-06 2:57:06
Message-ID: E1RMsuw-0005uZ-8G () bugs ! kde ! org
[Download RAW message or body]

https://bugs.kde.org/show_bug.cgi?id=279229


Parker Coates <parker.coates@kdemail.net> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|normal                      |wishlist




--- Comment #3 from Parker Coates <parker coates kdemail net>  2011-11-06 02:57:06 ---
Checking for "manual undos" seems silly. If the user knows that a move is going
to be a manual-undo, why don't they just hit Undo? If they don't know, then
they probably aren't expecting special behaviour. Adding extra code and further
complicating the solver to speed up such a corner case hardly seems like an
increase in efficiency.

Since May 2010, KPat the solvability is stored in the undo history, so that
particular request has been implemented. At that time, I experimented with
propagating solvability backward through the history and unsolvability forward
through the history. As you mention, it would theoretically reduce the time
spent processing and supply the user with better information. The problem is
that the solver isn't perfect; it produces both false positives and false
negatives. Trying to extrapolate from false information only compounds the
problem by increasing amount of false information. Maybe in the future, if the
solver's reliability has been much improved, we could look into this again.

-- 
Configure bugmail: https://bugs.kde.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching all bug changes.
[prev in list] [next in list] [prev in thread] [next in thread]