Reinhard, SMIRF in Gothic Chess Championship?

[Reinhard Scharnagl]

Hi Dan,

it would be something like a miracle, if I would be able to understand the time sceduling. Obviously the event is somehow late. But I am convinced that the GC team around Ed Trice will do its job.

If you would have read the texts on the GC games of Smirf you will have noticed, that Smirf already has been changed several times. The first attempt to fix a time management problem seems to work now. Then there has been a need to reduce the RAM requirements because the computer systems has been downgraded during the event to 256 MB RAM in total (as far as I know). Moreover I tried to improve Smirf after that, because Smirf has been "completed" in a hurry to make it able to participate. Thus Smirf has very week parts in its program except the basic data structure and the move generator.

There is a lot of optimizing potential within Smirf. If the actual Smirf would have started in that event as it is now, it would have gained more points. But the date has been part of the game ... so I hope for the last games ...

There is a cripelled Smirf beta shareware-version on Smirfs beta page: http://www.chessbox.de/beta.html .

Reinhard.

[Uri Blass]

No,
They added new loss to smirf in the results but did not fix the table correctly(they added a kiss for smirf in the table but the loss they forgot was not added).

Smirf now has 1.5/7 after it lost against the expected winner.

I expect it to win the last 6 games against weak opponents but it is only going to be enough for place 4 in the tournament.

I expect it to lose again against Ed trice so it is going to finish with 7.5/14

Uri

[Dan Honeycutt]

Hi Reinhard and Uri

It's Saturday evening in King of Prussia so the tournament should be finishing up. A couple games from Thursday have been added, including a loss by smirf. This gives it 1.5/7 as Uri says though the standings say 1.5/6. All those who gave Omid a hard time about slow updates during the WCCC should be following this tournament.

Reinhard, you had to reduce RAM requirements because they downgraded the hardware during the event? Are you serious? Wow.

I wouldn't worry about smirf's results. It has made its debut and I'm sure it will improve as time goes on. Good luck.

Dan H.

[Reinhard Scharnagl]

Hi Dan,

my RAM settigs had been 80, 160, 320 MB ... . During the development I have seen, that I have wasted space in each TT entry. When being informed, that 128 MB should be the absolute maximum of RAM used at the changed environment, I decided to do the adaption down to 64, 128, 256 MB ... , because I had planned that in each case. But the program has to be changed in anyway, because each day I have detected new weaknesses.

Smirf was improved during the event a lot, but changes still do not pay, except from some friendly comments of Ed Trice. :-)

Smirf prototype snapshot:
(German/English interface)

You are right: this has been Smirf's debut, and it's strength still has to be compared to other programs. Because its engine is also able to play 8x8 chess, there could be comparing games with other engines. But be aware that Smirf is still not using any opening libraries or table bases (as of course are implemented e.g. in Gothic Vortex).

Reinhard Scharnagl