The link to Feuerstein on your page seems to be broken. "Die von Ihnen aufgerufene Seite wurde nicht gefunden", my browser says.
While I am talking to you, let me once more express my gratitude for the work you do in testing UCI engines of all strengths. There is no other rating list which provides an estimate of the strength of a new version of a UCI engine as rapidly as yours. Whenever I release a new version of my engine (like now!), I press the "refresh" button on your site a hundred times per day until I see the first results. I am sure lots of other programmers do the same.
While I am talking to you, let me once more express my gratitude for the work you do in testing UCI engines of all strengths. There is no other rating list which provides an estimate of the strength of a new version of a UCI engine as rapidly as yours.
Tord
I'd like to echo my agreement. Keep up the great work Patrick.
The link to Feuerstein on your page seems to be broken. "Die von Ihnen aufgerufene Seite wurde nicht gefunden", my browser says.
Hi Tord,
The link is broken. I hope is not for a long time. Two days ago it works
Whenever I release a new version of my engine (like now!), I press the "refresh" button on your site a hundred times per day until I see the first results.
Tord
You must wait some days/hours
Now the 19th game has begun. Glaurung 2 - epsilon/2 (what a name!!) score 16/18.
There are 126 games in complet gauntlet. I think in 24h the first rating is out
Le Fou numerique wrote:The link is broken. I hope is not for a long time. Two days ago it works
OK, I hope it will work again soon.
Glaurung 2 - epsilon/2 (what a name!!)
Well, I've seen other programmers use version numbers like "1.99" for beta versions. As a mathematician, I found, "2-epsilon" both more elegant (why should the difference be exactly 0.01?) and more flexible (because I can continue with epsilon/2, epsilon/3, ... as long as I want).
On the other hand, it seems that lots of people miss the point and drop the minus sign, thus making the version number incomprehensible and stupid. Perhaps it wasn't a good idea after all.
Le Fou numerique wrote:The link is broken. I hope is not for a long time. Two days ago it works
OK, I hope it will work again soon.
Glaurung 2 - epsilon/2 (what a name!!)
Well, I've seen other programmers use version numbers like "1.99" for beta versions. As a mathematician, I found, "2-epsilon" both more elegant (why should the difference be exactly 0.01?) and more flexible (because I can continue with epsilon/2, epsilon/3, ... as long as I want).
On the other hand, it seems that lots of people miss the point and drop the minus sign, thus making the version number incomprehensible and stupid. Perhaps it wasn't a good idea after all.
I think in 24h the first rating is out
Just 24h? Very fast, as always.
Tord
Flexibility is no problem with 1.99 because you can continue from 1.99 to 1.991 and 1.992 and 1.999 and 1.9991 if you want.
Note that I have mathematical background and I had no problem to understand the point.
I thought that maybe I should add delta inside the name of movei because you know that for every epsilon>0 there is delta>0...
but I think that as long as movei is weaker than glaurung it is not a good idea to use delta inside the name of it.
Uri Blass wrote:Flexibility is no problem with 1.99 because you can continue from 1.99 to 1.991 and 1.992 and 1.999 and 1.9991 if you want.
Yes, you can, but it seems a lot more arbitrary than 2-epsilon, 2-epsilon/2, and so on.
I thought that maybe I should add delta inside the name of movei because you know that for every epsilon>0 there is delta>0...
but I think that as long as movei is weaker than glaurung it is not a good idea to use delta inside the name of it.
I'm not so sure - based on the latest results from Jorge, there doesn't seem to be a big difference in strength between our engines.
Feuerstein 0.4.5.1 first rating in the UEL: 2100.
