2
I Use This!
Inactive

Commits : Listings

Analyzed 1 day ago. based on code collected 1 day ago.
May 28, 2024 — May 28, 2025
Commit Message Contributor Files Modified Lines Added Lines Removed Code Location Date
Merge of generic games branch into trunk More... over 17 years ago
Changed start-up output to be a little bit better. If only the unit test line could be printed only when unit tests are actually runwq More... over 17 years ago
Added unit tests to random.d to test uniformity of random number generation More... over 17 years ago
Small updates to the last commit More... over 17 years ago
Yet another time upgrade More... over 17 years ago
An exhaustive replacement of exception output printing. This should help shed light on Urban's gdc execution issues. More... over 17 years ago
Fixed a huge mistake in the time handling... More... over 17 years ago
Yet another timing update. Timing parameter estimates get inaccurate as time gets short and the timing gets very sensitive to the adapative parameter changes. Instead of allowing crazy stuff to occur, this tries to fix the estimated lag parameters and trust the other adaptive features to adjust time usage in the final few moves of the game More... over 17 years ago
Makefile for raw use of gdmd More... over 17 years ago
Fix for gdc output flushing issues? More... over 17 years ago
Minor update to error reporting from gtp More... over 17 years ago
More updates for the time handling. Put a true lower bound on time per move (I hope in those cases it's lag estimation error). Also updated search to not resign without sufficient random sims to make decision. More... over 17 years ago
Updated time controls after losing many games on time on CGOS More... over 17 years ago
Added an extra benchmark to default output - number of random games played out since previous move. Should be useful for a quick check of speed improvements or looking at how fast one search algorithm is vs. another. More... over 17 years ago
Typo in make clean
Urban Hafner
as urbanhafner
More... over 17 years ago
Upgraded AMAF to use 5/8 of the moves in the random game (based on Don Dailey's guidance that this is near-optimal for this approach) More... over 17 years ago
A basic 'All Moves As First' MC variant More... over 17 years ago
Revised basic timing/thinking parameters. This should represent better defaults, and perform better on CGOS. More... over 17 years ago
Array initialization was in the wrong place. Previous defaults were acceptable, but with floats, it wasn't More... over 17 years ago
Updated shuffle mode operation based off revised theory of how to do it right. Will test on CGOS in the next few days... More... over 17 years ago
Fix of HUGE bug in shuffle-mode logic. Before fixing this bug, move selection was nearly random! More... over 17 years ago
Possible fix for bug 1779907 (crashes when running on CGOS) More... over 17 years ago
A small omission from the last commit More... over 17 years ago
Two small but important changes: More... over 17 years ago
A small help file that was accidentally omited from an earlier commit More... over 17 years ago
Updated makefiles More... almost 18 years ago
dsss_objs shouldn't be in svn as it gets deleted by dsss and when that
Urban Hafner
as urbanhafner
More... almost 18 years ago
Removed phobos support from tangoBind. Makes compilation with gdc simpler. More... almost 18 years ago
More reformatting of error display More... almost 18 years ago
Fixed problems associated with a short shuffle list. More... almost 18 years ago