"SMIRF".......Snifff.......Coff coff

[fern]

The downloadable engine available in the site is extremely weak. I would like to see another, stronger engine made with the techniques the author explain, more or less clearly.
An engine must be very very weak to be beaten by me at just 15 minutes the game because I am a lousy fast player, but this time I had a good time crushing the poor thing.

It is even weaker than CC7 regards
fernando

[nuff]

??? Which engine/site my good friend? What are you talking about? Can you give us a link please?

[smrf]

Well, SMIRF playing traditional chess is not yet a strong opponent like professinal programs, e.g. because it still is playing without using any opening library. Thus it is easy to experience an opening combination then outplaying SMIRF.

On SMIRF: traditional chess currently might be experienced as boring, if you are interested rather in creatively playing instead of reproducing moves out of your memory. Thus the main intention of SMIRF actually is to support random starting array based playing as necessary in Chess960.

Also there is only one small universal engine playing all variants, 10x8 chess included. 10x8 Capablanca Random Chess probably would be a real challenge for you to experience.

The free donationware version is intended for to introduce SMIRF's GUI and to unrestrictedly allow to investigate all its abilities including storing and loading games of all its supported variants (also traditional chess) using PGN files. The supporters' bonus engine is stronger, but not a world champion, even if it currently seems to be the strongest existing 10x8 chess program.

SMIRF is an amateur program not based on published sources. Its methods are experimental. SMIRF is about to be rewritten as a Mac OS X application, because it is my first finished chess program currently including some repairs, thus urgently needing a redesign.

Because of having to write an own universal GUI, SMIRF's engine could make good use of putting more time into that instead of its GUI.

http://www.chessbox.de/Compu/schachsmirf.html

Reinhard Scharnagl.

[Uri Blass]

The downloadable engine available in the site is extremely weak. I would like to see another, stronger engine made with the techniques the author explain, more or less clearly.
An engine must be very very weak to be beaten by me at just 15 minutes the game because I am a lousy fast player, but this time I had a good time crushing the poor thing.

It is even weaker than CC7 regards
fernando

I did not play against smirf but I think that without seeing the relevant game I cannot say if the engine played bad or you played well.

I also think that a single game proves nothing because an engine may have some bug and you can be lucky to take advantage of the bug in your first try.

Uri

[fern]

I did not save the game, as I almost never do. But surely you will agree with me that a program that without reason play h5 and then move the rook to h4 just to be attacked by a bishop and so on, moves like that all around, well, I think it is enough.
I am strong club player, 2100 or so, but I can swear you this pogram such as it is the downloadabale version, can be beaten without problem by a 1300 player.
Nevertheless, I hope the new techniques will bear fruits sooner or later.

My best
Fernando

[fern]

The weakness I saw in Smirf has nothing to do with lack of an openning book, neither I made use of any prepared device of such kind to win. Simply Smirf plays moves -as black- as h5 without reason, then moves the rook along the column, then the rook come back, things like that. Pity I did not save the game. It was at 15 minutes and I mated it without problem, but anyway he has lost on time.
Perhaps that is creative, but the kind of it escapes my understading.
I am sure you will, at last, present something amazing. I hope so because I am somewhat fed up of the traditional engines, always winning me.
No joke: I really expect new, innovative engines.
If I bothered you, sorry.
I know first trys are always difficult.
Keep the good work.

My best
Fernando

[smrf]

Fernando, I prosume you have played a blitz traditional chess game against SMIRF. SMIRF is intended to play positionally. And for its author to become able to see progresses or bugs within that engine, it still is avoiding the use of opening libraries. Thus any experienced player applying his opening knowledge would for sure win, especially doing blitz timings.

For to experience SMIRF's abilities it either should be experimented by using test positions (no endgames yet please, because SMIRF does also not yet support table bases), or by playing Chess960 or CRC games. And, if being interested in its development, it might be a good idea to support SMIRF, and then to use the improved bonus engine.

Reinhard.

[fern]

http://www.chessbox.de/Compu/schachsmirf_e.html

[smrf]

Fernando, obviously you have been able to install, use and experience SMIRF's donationware release without any problem. But just that is, for what that introducing release has been intended for.

If that could raise your interest in SMIRF - fine. If not, of course, it's ok as well.

The SMIRF FullChess idea is to play several downwards compatible chess variants using a unique engine currently avoiding any table knowledge.

Reinhard.

[fern]

I am interested in any endeavor conneected with chess computers, so with yours.
I thinks perhaps it would be not a bad idea to include a book. Let smirf proves his worth in the middle game. Even seasoned commercial engines fall in problems playing without it, so to compel smirf to that test is perhaps too much.

My best
Fernando