My configure fails with
checking for python extension module directory...
${exec_prefix}/lib/python2.5/site-packages
./configure: line 22916: syntax error near unexpected token `have_python="yes",'
./configure: line 22916: `AM_CHECK_PYTHON_HEADERS(have_python="yes",
have_python="no")'
Any ideas?
I did all my development in python. Look at gnubg-nn and see if there
is anything you like. The problem is that it uses my own cube code
(which I think is as good or better than gnubg's)
-Joseph
On Fri, Jan 1, 2010 at 10:33 AM, Mark Higgins wrote:
> I'm coding up my own little backgammon simulat
Hi,
I think it is a shame gnubg is not represented in the mobile/tablet world.
I know next to nothing about android, but from my understanding you
can develop in Java, or even Python, and call C or C++ for speed
(neural net ...)
Does anyone (or maybe a friend) would like to collaborate and give
I can't really remember. It is entirely reasonable to assume that with
the increase in CPU power we can have a better soft spot of prunning
net size and number of positions to prune.
-Joseph
On 17 February 2011 10:50, Philippe Michel wrote:
> I have tried to use the patch to SSE code that Oystei
Hi Michael,
Can you please post the detailed instructions of how to build GNUBG
for the latest machines?
I have obviously built GNUBG many times before, but last time I
suspect the build was not optimal, and uses only a single core.
So I am specifically referring to the flags to give ./configure
A far more interesting and difficult task would be to go for one of
the other variants In my family they used to play machboosa
(surrounded), which is the same as the Greek Plakoto, which I find
strategically orders of magnitude harder. I can hardly tell if I make
sensible moves or not, as this is
he
> publications section.
>
> Cheers,
> Nikos Papahristou
>
> On 28/6/2011 23:17, Joseph Heled wrote:
>>
>> A far more interesting and difficult task would be to go for one of
>> the other variants In my family they used to play machboosa
>> (surrounded),
This code (which uses random()) would have no bias and may even be faster,
bypassing floats
const long rng = RAND_MAX/6;
const long limit = rng * 6;
/* get one random dice */
int getd() {
long v = limit;
while( v >= limit ) {
v = random();
}
return v/rng; // dice in [0..5], add 1 fo
Well, I am not sure how you flip the position, since it matters who is
on the move.
-Joseph
On 10 December 2011 16:17, Mark Higgins wrote:
> I've been playing around a bit with neural networks for backgammon and found
> something interesting, and want to see whether this is already part of gnub
My experience tells me that 100,000 trials may not be sufficient.
With today's computing power it should be easy to do at least a
couple of millions.
-Joseph
On 12 December 2011 11:22, Mark Higgins wrote:
> I tried a little experiment on this: a 10-hidden-node network with a single
> probabil
It is true that for moves we are interested only in relative
evaluations between moves, but in cube actions we are interested in
relative and value against the drop point.
The GNUbg training data contained both "move" positions and "cube
action" positions.
My idea was (and I am not sure if this wa
I find it hard to believe.
There is a simple test - if the two implementations give the same vale for
the same position, ply 0.
-Joseph
On 21 December 2011 09:18, Øystein Schønning-Johansen wrote:
> On Tue, Dec 20, 2011 at 8:41 PM, Philippe Michel
> wrote:
>
>> In lib/sigmoid.h, there is some
ementation it was trained with.
In the past, breaking this meant a serious drop in move quality. I
suggest someone checks this
-Joseph
On 22 December 2011 02:03, Philippe Michel wrote:
> On Wed, 21 Dec 2011, Joseph Heled wrote:
>
>> I find it hard to believe.
>>
>> T
Many moons ago, I created bglight (http://savannah.nongnu.org/projects/bglight/)
It had the required separation as a design goal, but nothing came out of it.
I am still amazed that with today's power, there is no ipad/andriod
port of gnubg.
-Joseph
On 22 December 2011 23:11, Guido Flohr wrote:
Seems like you got a very very slightly better race net, but I would be
surprised if it makes a difference in real life.
Would be much more interesting to
- get a better contact or crashed net
- expand the roll-out database for all categories (should be easy with
the current availability of cy
(especially with regard to
the odd/even issue).
-Joseph
On 5 January 2012 21:17, Massimiliano Maini wrote:
> On 5 January 2012 02:20, Joseph Heled wrote:
> > Seems like you got a very very slightly better race net, but I would be
> > surprised if it makes a difference in real life.
&
t to R.Janowksi's formulae (and it was).
>
> Not sure I can help in integrating your code in gnubg, but I'm definitely
> curious in understanding what's different with respect to the current one.
>
> MaX.
>
> On 5 January 2012 09:35, Joseph Heled wrote:
> > I
On 6 January 2012 10:03, Philippe Michel wrote:
> On Thu, 5 Jan 2012, Joseph Heled wrote:
>
> Seems like you got a very very slightly better race net, but I would be
>> surprised if it makes a difference in real life.
>>
>
> Ok, so a worthwhile improvement would be a
The 0.17 weights and the benchmark files? They are no longer on
ftp://ftp.demon.nl/pub/games/gnubg/nn-training/benchmarks
Thanks, Joseph
___
Bug-gnubg mailing list
Bug-gnubg@gnu.org
https://lists.gnu.org/mailman/listinfo/bug-gnubg
My thinking was (and it was only my judgement, non tested, so could be
wrong) is that a benchmark consisting of computer only positions would
be more biased than one with some human input. Note that human input
can go both ways; first you get positions arising from bad plays, and
second you may get
Hi,
I just run 1 1 point match games between gnubg and pubeval. I
think this is the most fair, since there are no gammons of cubes
involved.
However, gnubg won 9840 of those, which makes me think something is wrong.
Can someone help me verify that pubeval makes the "right" moves?
I can send
t;
> http://www.bkgm.com/rgb/rgb.cgi?view+610
>
> ?
>
>
>
> On Jan 19, 2012, at 8:31 AM, Joseph Heled wrote:
>
>> Hi,
>>
>> I just run 1 1 point match games between gnubg and pubeval. I
>> think this is the most fair, since there are no gammons of cubes
&
t;
> On Jan 19, 2012, at 9:39 AM, Joseph Heled wrote:
>
>> I was asking for help, not for more work.
>>
>> For example, anyone who implemented play against pubeval can check a
>> few of the moves I sent.
>>
>> -Joseph
>>
>> On 20 January 2012 03:35,
gt; *|| | | | | ||| | | | | ||*
>> *|o *| * o|*
>> *|o *| * o|*
>> *| *| * o|*
>> *| *| o|*
>> *| *| o|*
>> *| | |*
>> *| | |*
>> *| | |*
>> *| o| *|*
>> *| o| *|*
>> *| o| *|*
>> *|* o| o *|*
>> *|* o o| o *|*
>>
basic net evaluation?
-Joseph
On 20 January 2012 12:03, Nikos Papahristou wrote:
> Interesting.
> For comparison reasons with Palamedes and TD-Gammon, backgammons should
> count as gammons. This changes the second performance metric to 0.6046ppg.
>
> Nikos
>
> On 20/1/2
-2620M CPU @ 2.70GHz
-Joseph
On 20 January 2012 12:03, Nikos Papahristou wrote:
> Interesting.
> For comparison reasons with Palamedes and TD-Gammon, backgammons should
> count as gammons. This changes the second performance metric to 0.6046ppg.
>
> Nikos
>
> On 20/1/2012 0:20 π
CRASHED attempts to capture the positions where one side has only a
small number of "active pieces". The number of active pieces has been
arbitrarily set at 6, and the definition requires that you have at
most 6 checkers not on points 1 or 2, accounting for the possibility
of one checker from 2 se
On 9 February 2012 20:33, Michael Petch wrote:
> On 09/02/2012 12:01 AM, Mark Higgins wrote:
>> Can anyone point me to a function to convert the keys in the benchmark dbs
>> to a board layout pls?
>>
>> They don't look like 14-char position IDs, and checking this list archive I
>> see them occas
'r' is the seed used for the rollout, I think
On 10 February 2012 23:02, Øystein Schønning-Johansen
wrote:
> 2012/2/10 Mark Higgins
>>
>> Thx - what is the contact.bm format then? It didn't seem as
>> straightforward as the training one.
>>
>
> A bit more complex...
>
> lines starting with 's',
if the bot does not care about points lost, it should resign, no?
-Joseph
On 15 February 2012 11:23, Michael Petch wrote:
> As an addendum. Bots often make what seem like unnatural plays compared
> with that of a human. If every decision has the same equity the bot will
> use the one that happen
Apply my argument again:) The bot should roll, then resign.
(I am not saying how hard or easy it is to do in gnubg, but just what
makes sense for a bot)
-Joseph
On 15 February 2012 11:47, Michael Petch wrote:
> On 14/02/2012 3:32 PM, Joseph Heled wrote:
>> if the bot does not care abo
I was not suggesting any changes. I was merely pointing out what I
would find the best behaviour from a bot. Others may disagree, or
dispute the degree of annoyance to the 'user'
-Joseph
On 15 February 2012 12:11, Russ Allbery wrote:
> Michael Petch writes:
>
>> If I were Russ, I wouldn't be bo
Converting position probabilities into cubeful error measures is open
to personal interpretation.
My script (perr.py) provides measures for a variety of match scores [
(7,7), (3,3), (5,5), (9,9), (25,25), (2,3), (2,4), (2,5), (2,6),
(2,7), (3,4), (3,5), (3,6), (3,7), ... ]
Note: The cube decision
Hi,
This position just falls under the special code I wrote ages ago.
There might be errors (bugs) there, but please remember that the
"exact" is *not* a '8 vs 2 1 0 0 3 cubeless bearoff', since clearing
checkers from the back is not the same as bearoff - you *can* move any
checker you want. I t
On 27 February 2012 19:09, Philippe Michel wrote:
> On Mon, 27 Feb 2012, Joseph Heled wrote:
>
>> clearing checkers from the back is not the same as bearoff - you *can*
>> move any checker you want.
>
>
> Embarrassing. You're right of course and in my example 3s
100.00% 100.00%
1 ply: 1.000 1.000 0.678 0.000 0.000 100.00% 100.00%
2 ply: 1.000 1.000 0.799 0.000 0.000 100.00% 100.00%
On 27 February 2012 19:33, Joseph Heled wrote:
> On 27 February 2012 19:09, Philippe Michel wrote:
>> On Mon, 27 Feb 2012, Joseph He
%
1 ply: 1.000 1.000 0.678 0.000 0.00067.61%67.61%
2 ply: 1.000 1.000 0.681 0.000 0.00067.64%67.64%
On 27 February 2012 20:46, Joseph Heled wrote:
> Not sure how to set up gnubg to evaluate higher plies. This is after
> manually editing the position
>
&
On 19 June 2012 09:36, Philippe Michel wrote:
> On Tue, 5 Jun 2012, Philippe Michel wrote:
>
> The benchmark database for the crashed positions seems seriously
>> corrupted.
>>
>
> I have rerolled it. How should I proceed to have it uploaded to
> ftp.demon.nl ?
>
I should remind everyone that t
Hi,
It is heart warming to see that the race net can be improved still :)
The benchmark rates the new crashed net as stronger, which is great too,
but the new contact net is rated weaker in moves (0.0104453134853 vs
0.0104996763558), stronger in cube actions.
-Joseph
On 22 June 2012 09:37, Phil
On 24 June 2012 13:38, Joseph Heled wrote:
> Hi,
>
> It is heart warming to see that the race net can be improved still :)
>
> The benchmark rates the new crashed net as stronger, which is great too,
> but the new contact net is rated weaker in moves (0.0104453134853 vs
On 25 June 2012 04:13, Philippe Michel wrote:
> On Sun, 24 Jun 2012, Joseph Heled wrote:
>
> On 24 June 2012 13:38, Joseph Heled wrote:
>>
>> Hi,
>>>
>>> It is heart warming to see that the race net can be improved still :)
>>>
>>> The
Can we open a shared dropbox for those files? or gdrive on a google account?
-Joseph
On 25 June 2012 08:42, Philippe Michel wrote:
> On Sun, 24 Jun 2012, Jim Segrave wrote:
>
> Can someone supply me with a list of the files to be updated and where I
>> can get the new versions?
>>
>
> I had se
Hi again Philippe,
Did you find a way to show that the new net indeed is indeed more balanced
than the old with regard to the odd-even ply syndrome?
-Joseph
On 25 June 2012 07:37, Philippe Michel wrote:
> On Sun, 24 Jun 2012, Joseph Heled wrote:
>
> I am very interested to know
Hi again,
While the benchmark shows the new net(s) are better than the old, one must
remember the improvement in playing ability is not large.
In 1 million 5 point matches between the two, the new net wins in 50.084%
of the matches. This result is not even strongly significant, at about 2
sigma.
Hi,
While "importing" the sse code to gnubg-nn I discovered an error in the
sigmoid implementation. The calculation made by the see code did not match
the one from the straight C code. (It was relatively easy for me to catch
since sse can be toggled during run time in gnubg-nn)
Since this results
On 3 July 2012 08:34, Philippe Michel wrote:
> On Mon, 2 Jul 2012, Joseph Heled wrote:
>
> While "importing" the sse code to gnubg-nn I discovered an error in the
>> sigmoid implementation. The calculation made by the see code did not match
>> the one fr
Hi
What did you use to play those 1M games?
-Joseph
On 18 September 2012 10:54, Philippe Michel wrote:
> I have uploaded new weights at
> http://www.gnubg.org/media/nn-training/pmichel/nets/20120907/
>
> The difference since what I posted in June is mostly further training of the
> contact net,
statistically significant?
-Joseph
On 18 September 2012 18:51, Philippe Michel wrote:
> On Tue, 18 Sep 2012, Joseph Heled wrote:
>
>> What did you use to play those 1M games?
>
>
> This was with your gnubg-nn/scripts/play/matchplay.py script.
__
when running the matches, did you use either -q or --q1 on the command line ??
-Joseph
On 19 September 2012 08:10, Philippe Michel wrote:
> On Tue, 18 Sep 2012, Joseph Heled wrote:
>
>> Just in case you are unaware of it, my code does not reflect gnubg
>> playing ability per
trial, -,0,+ (X
winds both, they split, O wins both), so the stats will be different.
But I would repeat with a -q1. 50.69% is huge. I don't expect it to go away.
-Joseph
On 19 September 2012 09:05, Philippe Michel wrote:
> On Wed, 19 Sep 2012, Joseph Heled wrote:
>
>> when ru
your test statistic is 117*-1 + 262*0 + 121*1 = 4
the standard deviation is (500 * (2/3))^0.5 = 18.25
So, far from significant :)
If 50.6% is correct, you should get significance with 200,000 matches.
Should be easy with 1ply
-Joseph
On 19 September 2012 18:45, Philippe Michel wrote:
> 262 121
(OO) = 1/4, P(XO) =
1/2 the mean is zero and the variance is 1/4*1^2 + 1/4*(-1)^2 = 1/2.
The variance of N matches is N* (1/2) and so the std is (N/2)^0.5
-Joseph
On 20 September 2012 08:35, Philippe Michel wrote:
> On Wed, 19 Sep 2012, Joseph Heled wrote:
>
>> your test statistic is
Hi,
I am in favour of
-- uniform style in a project
-- not mixing tabs and spaces. I would say use spaces only.
-- reasonably cramming as much code into HxW as possible : my
eyesight forces me to relatively large font, even on a large screen.
So I *hate* lines with a single curly (wasti
in the early- to mid-2000s when the currents nets were trained is
> now doable.
>
> I don't know if Joseph Heled did many iterations (reevaluate database /
> train nets / maybe add mishandled positions) but with rollouts, each of
> them take a long time (I did it twice for the c
In my book this (Boland) is trolling and I suggest you stop feeding it.
-Joseph
On 13 March 2013 16:08, Russ Allbery wrote:
> Michael Petch writes:
>
> > There isn't anything wrong with the dice, and the computer isn't getting
> > any luckier than it would in real life with properly weighted
With all those nets I get confused.
Which are exactly the "old" in "new" (in a nngnu ready format too), and
more important, where can I download them now.
In my gnubg directory I see
$ md5sum gnubg.weights*
d182a162a011aac839bc48879c349b14 gnubg.weights
bb76fbaa6256f79dacc17749045d38ee gnubg.w
Michel wrote:
> On Mon, 18 Mar 2013, Joseph Heled wrote:
>
> With all those nets I get confused.
>>
>> Which are exactly the "old" in "new" (in a nngnu ready format too), and
>> more important, where can I download them now.
>>
>
> &quo
Any ideas what went wrong ?? I am on ubuntu linux, not Darwin or Windows
./autogen.sh
+ rm -f aclocal.m4
+ aclocal -I m4
+ uname -s
+ [ Linux = Darwin ]
+ LIBTOOLIZE=libtoolize
+ rm -f m4/libtool.m4 m4/ltoptions.m4 m4/ltsugar.m4 m4/ltversion.m4
m4/lt~obsolete.m4
+ libtoolize --force --copy
libtool
Hi Rick,
It has been a long time. The mec26 was obtained by first getting the post
crawford probabilities from rollouts, then from the program mec.cc (i.e.
"first principles"). I can't remember where the other tables came from :(
However, I remember running a fair number of rollouts and evaluatio
It is due to a special trick wrist we put inside computer, which knows how
to roll doubles.
-Joseph
On 24 January 2014 20:39, Viorel Crt wrote:
> in each game i play computer have more doble, usually doble than me
>
> ___
> Bug-gnubg mailing list
> B
On 30 January 2014 01:57, Øystein Schønning-Johansen wrote:
> Hi!
>
> A question mainly to Joseph: Did you ever run any move benchmarks on
> the move selection algorithm for the one-side-rollout code?
>
Can't remember details, but while the OSR was great for lifting the race
net off the ground, t
You may want to look again at gnubg-nn inputs.c . It does some pre
computing (into BoardCache) which is used by the various individual
inputs. I think this takes care of most repeated computations. Also,
gnubg-nn uses my modified eggmove for generating moves which is far faster
than gnubg main co
Why didn't the opponent resign before? I think it is a just punishment :)
-Joseph
On 2 August 2014 05:51, Michael Petch wrote:
>
>
> On 2014-08-01 11:29 PM, Michael Petch wrote:
> > What is going on is that both plays one off or 2 off still guarantee
> > victory. The bot sees that it has a cho
Hi,
I upgraded to a 64 bit kubuntu mint 17, and make fails rather strangely.
Anyone knows what is going on?
> make
/bin/bash ./ylwrap sgf_y.y y.tab.c sgf_y.c y.tab.h `echo sgf_y.c | sed -e
s/cc$/hh/ -e s/cpp$/hpp/ -e s/cxx$/hxx/ -e s/c++$/h++/ -e s/c$/h/` y.output
sgf_y.output -- yacc -d
/home/j
Well, I would love it if the display engine could be bundled with a FIBS
client sans the brainz.
I stopped playing on FIBS for some years now due to lack of a Linux client.
-Joseph
On 28 June 2015 at 23:55, Lucas wrote:
> Hi
>
> Gnubg is open source.
> This means when the add code to use gnu
2015 at 15:42, Michael Petch wrote:
>
>
> On 2015-06-28 9:30 PM, Joseph Heled wrote:
> > Well, I would love it if the display engine could be bundled with a FIBS
> > client sans the brainz.
> >
> > I stopped playing on FIBS for some years now due to lack of a Linu
s. I found Oystein Johansen's paper from
> December 2007 on the GNUBG website. It credits Joseph Heled with the main
> ideas, and points to his website, but the only links I could find to
> Heled's website were all dead. Is there any way I can obtain more
> documentation about wh
Hard for me to see how this can be used, since it is not "Data" in the
usual sense.
That said, I have no objections.
-Joseph
On 13 October 2016 at 10:09, Øystein Schønning-Johansen
wrote:
> Philippe wrote:
> *I'm not sure how the could get something out of that in the current
> format, with th
s. This is probably the most
> promising way to improve the level of play, but this seems to be pretty
> hard. The inputs currently used are mostly (or even exclusively) from old
> research papers by Hans Berliner 40 years ago. I don't think there is a lot
> more recent litterature on
Probably many known ways to convince people there is not cheating. for
example
1. Before the math, GNUBG outputs an encrypted file with all "his" dice in
a sequence.
2. Match played. crybaby supplies dice (GNUBG adds his "dice" mod 6) to get
actual dice used.
3. After the match, GNUBG gives crybab
Since backgammon pieces don't come back from the dead, a system with
plentiful memory will have a set of nets for
one-side-has-1-checker, one-side-has-2-checkers, etc/ That will help a lots
of those boundary (but relatively rare) situations.
Another possible approach in that direction is to have a
Here is another fun project. Solve BG for 1 checker vs 1, exactly.
Obviously, the only interesting case is when you have to decide if you want
to hit or not.
On 18 February 2018 at 12:21, Joseph Heled wrote:
> Since backgammon pieces don't come back from the dead, a system with
>
I apologize profusely as this is technically a non-backgammon post. But no
doubt some of you will find it interesting, and anyway I am soliciting help
:)
I just uploaded a new project to github called royalur (
https://github.com/jheled/royalur). The documentation is at
https://jheled.github.io/ro
Don't know if it will work, but have you tried 8 W pieces on point 2,
one black piece on 3 and the rest black pieces at B home. W can win
with two doubles (2 to 6), or lose with a sequence of (2-1, 3-1,...)
throws while B gets a sequence of 6-6. The odds will be low but might
be > 0.
-Joseph
On W
Hi,
(I know this is the wrong forum, but casting the net anyway)
I am looking for someone who might want to collaborate on a project of mine
about racing games .
(It is a "fun" project, like gnubg, by which I mean non-paid and done out
of "love". I can't pay anyone since I am not getting paid my
You have many independent runs, right? Why worry about multi-threading?
Divide the set into (say) 16 threads (or whatever makes sense for your CPU)
and run each set on another thread.
-Joseph
On Mon, 17 Jun 2019 at 19:28, Øystein Schønning-Johansen
wrote:
> Thanks for your input. I'm still thin
And don't use OSR. This was good for bootstrapping in the old days, when
CPU was much slower.
On Mon, 17 Jun 2019 at 19:33, Joseph Heled wrote:
> You have many independent runs, right? Why worry about multi-threading?
> Divide the set into (say) 16 threads (or whatever makes sense f
A wonderful initiative. Perhaps one day we will have a reasonable Android
version and all will be well.
-Joseph
On Sun, 22 Sep 2019 at 19:00, Theodore Hwa wrote:
> Hi gnubg users,
>
> I have ported GNU Backgammon to the web, replacing the GTK UI with a
> Javascript UI. You can see it hosted her
I personally would not spend time on this minor issue at this point. If you
want to save space you can try and train dedicated nets, which might
outperform the one-sided approach.
-Joseph
On Wed, 25 Sep 2019 at 08:37, Philippe Michel
wrote:
>
> On Sat, Sep 21, 2019 at 07:15:05PM -0700, Theodore
A nice 2d Board will be more that useful...
On Thu, 26 Sep 2019 at 22:03, Jon Kinsey wrote:
> The OpenGL 3d board is theoretically portable to WebGL. In practice this
> is likely to be difficult... I’ll take a look, but I have little free time
> so will be a while before I report back on my init
Hi Ian,
You obviously put a lot of thought and effort into this project. Here are
some initial reactions (more might come later).
1. RE rollouts, you probably want to use the Python interface.
Unfortunately I have been out of the loop for a long time, and can't
remember the details.
2. I suggest
Hi Timothy,
Here is a stats question I encounter from time to time.
Suppose I run N BG games and collect the average win rates and gammon
rates. 4 estimates which are dependent as they sum to 1.
How do I determine the confidence intervals for each? This is a 4d vector
and it seems like a non triv
ves things
like contour integration or other horrors. I hoped for something simpler.
-Joseph
On Wed, 13 Nov 2019 at 06:04, Timothy Y. Chow
wrote:
> On Tue, 12 Nov 2019, Joseph Heled wrote:
> > Hi Timothy,
> > Here is a stats question I encounter from time to time.
> >
&
I thought it was clear that what we want to establish a difference between
two (say players X and Y) by running games and testing that gammon-rate(X)
!= gammon-rate(Y).
-Joseph
On Wed, 13 Nov 2019 at 10:39, Timothy Y. Chow
wrote:
> On Wed, 13 Nov 2019, Joseph Heled wrote:
> > &quo
wrong and it is the other
way. To get a weaker player you probably need to use a "weaker/lesser
trained net".
-Joseph
> Cheers, Ian.
> On Wednesday, 27 November 2019, 05:59:09 am AEDT, Joseph Heled
> wrote:
>
>
> Hi Ian and bg fans,
>
> I would like to clarify my ow
>
> Another option might be to use a very old build of Gnubg. There are some
> significant strength gains along the way to our current version. You did the
> neural net, you would know that better than anyone :-)
>
> On Wednesday, 27 November 2019, 11:38:17 am AEDT, Joseph Heled
Anyone knows where the files are? Perhaps we can host them somewhere
else? I might have some space on my personal cloud.
-Joseph
On Thu, 28 Nov 2019 at 00:58, Thomas A. Moulton wrote:
>
> This could be part of the problem...
>
> http://gnubg.org says
>
> You will not see anything here anymore.
>
Thank you. Looks good enough for me.
-Joseph
On Thu, 28 Nov 2019 at 20:39, Philippe Michel wrote:
>
> On Wed, Nov 27, 2019 at 10:30:38AM +0100, Øystein Schønning-Johansen wrote:
>
> > I'm trying to download the training data from the site.
> > http://files.gnubg.org/media/nn-training/
> >
> > Se
I am out of the loop too, but speeding up rollouts (i.e. using modern
multicores) seems like a worthy improvement.
-Joseph
On Wed, 4 Dec 2019 at 07:08, Øystein Schønning-Johansen
wrote:
>
> Hi Sarah!
>
> Thanks for taking contact. Good to hear that you like GNU Backgammon.
> Is it still under d
)
-Joseph
On Wed, 4 Dec 2019 at 09:38, Øystein Schønning-Johansen
wrote:
>
> On Tue, Dec 3, 2019 at 7:37 PM Joseph Heled wrote:
>>
>> I am out of the loop too, but speeding up rollouts (i.e. using modern
>> multicores) seems like a worthy improvement.
>
>
> Isn
said. Does modern processors come with AVX512 these days? Do we have
> support for that? In it is common to have AVX512 and we don't have support
> for it, maybe that is something that can go on the TODO list?
>
> -Øystein
>
> On Tue, Dec 3, 2019 at 10:07 PM Joseph Heled w
rce or install a binary from a packet manager?
> Does it say multi-thread support in: Help->About->Build Info ?
>
> -Øystein
>
> On Tue, Dec 3, 2019 at 9:46 PM Joseph Heled wrote:
>>
>> All I know is that I installed gnubg from the repository (GNU
>> Backgammon 1.0
*only*
in rollouts and so everything else works on older systems?
Is that possible or am I just showing my ignorance?
-Joseph
On Wed, 4 Dec 2019 at 10:42, Russ Allbery wrote:
>
> Joseph Heled writes:
>
> > My cpu (i7) has sse4 and avx2 support. Compiling from source is not the
&g
-- Forwarded message -
From: Joseph Heled
Date: Wed, 4 Dec 2019 at 11:07
Subject: Re: current development
To: Russ Allbery
Yes, I assumed GNUBG is not ready yet, but was not sure if you can
actually compile a (say) version with SSE2 and AVX support which will
run as long as the
Just for the record: 450924 evaluations/second.
-Joseph
On Wed, 4 Dec 2019 at 12:49, Joseph Heled wrote:
>
> I set number of threads to 4
>
> Rolled a cube decision
>
> world-class: 2-ply for move/cube-action : 18 minutes
> expert (0ply) moves/ 2-ply cube actions: 2 mi
d, Dec 04, 2019 at 09:46:09AM +1300, Joseph Heled wrote:
>
> > All I know is that I installed gnubg from the repository (GNU
> > Backgammon 1.06.002) and the rollouts speed seemed terrible.
> >
> > Do I need to do something in the GUI to enable multi-threading? (and I
> &
tried to compile gnubg
missing glib2
any ideas?
checking for GLIB... no
configure: error: You need to have glib2 version 2.8.0 or higher to
compile GNU Backgammon
joseph@blkdow:~/Projects/gnubg/gnubg$ sudo apt-get install glib2-devel
Reading package lists... Done
Building dependency tree
Reading s
Is that the right way to specify both?
./configure --enable-simd=avx --enable-simd=sse2
Thanks, Joseph
On Wed, 4 Dec 2019 at 13:13, Joseph Heled wrote:
2 -DG_DISABLE_ASSERT
-MT libboard3d_la-DrawOGL.lo -MD -MP -MF .deps/libboard3d_la-D
rawOGL.Tpo -c DrawOGL.c -o libboard3d_la-DrawOGL.o
DrawOGL.c:25:10: fatal error: Shapes.inc: No such file or directory
#include "Shapes.inc"
^~~~
On Wed, 4 Dec 2019 at 13:21, Joseph Heled w
bg to deteriorate that much. Pretty sad.
-Joseph
On Wed, 4 Dec 2019 at 13:24, Joseph Heled wrote:
>
> It is the usual hell of installing all missing dev packages and
> getting configure to complete.
>
> But this is unexpected: DrawOGL.c:25:10: fatal error: Shapes.inc: No
>
1 - 100 of 296 matches
Mail list logo
