New Little Goliath 3.15 with NN released

[Modern Times]

The developer *owes* us nothing but *gifted* us something special.

Exactly right. And if anyone doesn't want to use it for a particular reason, that is just fine also.

[carldaman]

I'm sorry but I feel like I'm being fooled. If someone who has been programming for many years is still only offering single core engines in 2022, that's poor. It's ok for new developers, but those who program engines for longer should show MP versions.

It's up to the author what kind of things he want to do. Michael is experienced with other neural nets, and didn't know about this new technics in computerchess. After 15 years away from chess he found out about nnue, wanted to try it with LG, and offers it for free. You feel fooled because he didn't first implement multi cores for you? Just leave it, if you don't like it. I like it, and I am really excited to play LG NN!

I will write my opinion. The programmers should know that not everyone cheers as soon as something new is introduced here. In 2022, I expect programmers to adapt to reality as well. The fact is that there are no single core computers for sale, even my 11 year old laptop has 2 cores (4 threads). I can't do much with single core engines. I'm sorry, but I would like to be able to express my opinion. And yes, I have Little Goliath installed, it does 5000 kns on a core (wow), but that's pretty much all this engine can do.

Dude, not *everyone* is trying to edge out the latest CPUs on PlayChess and so needs powerful hardware.

To me, LG is an absolute CLASSIC engine (I'm in my late 30s now and loved Little Goliath Light 1.5 even as a kid in 2002 in my Fritz 7 GUI) and its playing style with NN is very interesting. In contrast to the old attacking Goliath, it seems to be pretty positional.

It plays like no other engine if you observe its games.

This unique "opinion" of LG is definitely useful for analysis for humans, even with a single CPU.

The developer *owes* us nothing but *gifted* us something special.

It's "looking a gift horse in the mouth". Demands instead of gratitude. If it were a commercial product, it would be understandable.

[Alexander Schmidt]

I expect programmers to adapt to reality as well.

There are thousands of engines out there. If one doesn't fulfill your expectations, just leave it.

This kind of attitude rose in the last year in this community. The engines are for free, and the authors can do what they want. And I bet there would be much more nice engines, if programmers wouldn't have to read this kind of stuff all the time. I know several programmers who quitted because of this.

You would have a much better chance for a LG MP version if you would say something like: "Thank you very much, do you plan to implement MP? Would be great!" Instead of "You fooled me"

Michael just registered here, I hope he didn't immediately lose his returned desire for computerchess... He works with NN and maybe could bring new ideas to computerchess.

[Eduard]

I don't criticize everyone. But I really don't understand that someone who has been programming for 25 years or more, can only present the user a single version in 2022. I remember the time when German chess newspapers advertised a commercial (!) Goliath chess program (not Little Goliath) as "coming soon". Didn't come. Anyone who was already thinking about the commercial side 20 years ago, should be able to create an MP version in 2022.

[Gabor Szots]

I don't criticize everyone. But I really don't understand that someone who has been programming for 25 years or more, can only present the user a single version in 2022. I remember the time when German chess newspapers advertised a commercial (!) Goliath chess program (not Little Goliath) as "coming soon". Didn't come. Anyone who was already thinking about the commercial side 20 years ago, should be able to create an MP version in 2022.

So in your opinion is it better not to have LG at all than have it without SMP? If you were a waiter you'd only accept big tips and reject small ones?

[Rebel]

I don't criticize everyone. But I really don't understand that someone who has been programming for 25 years or more, can only present the user a single version in 2022. I remember the time when German chess newspapers advertised a commercial (!) Goliath chess program (not Little Goliath) as "coming soon". Didn't come. Anyone who was already thinking about the commercial side 20 years ago, should be able to create an MP version in 2022.

What a nonsense.

What if a programmer doesn't care?

The fun and creativity is in the single version, adding MP is just boring technical stuff which has nothing to do with chess itself.

[Eduard]

The point is that someone here hasn't been able to create an MP version for more than 20 years. I'm disappointed about that.

[Eduard]

I don't criticize everyone. But I really don't understand that someone who has been programming for 25 years or more, can only present the user a single version in 2022. I remember the time when German chess newspapers advertised a commercial (!) Goliath chess program (not Little Goliath) as "coming soon". Didn't come. Anyone who was already thinking about the commercial side 20 years ago, should be able to create an MP version in 2022.

What a nonsense.

What if a programmer doesn't care?

The fun and creativity is in the single version, adding MP is just boring technical stuff which has nothing to do with chess itself.

Ed, unfortunately you're living in the past too. Take Shredder, Hiarcs, Fritz, etc. as an example. But you don't have to. But don't be surprised that I'm no longer interested in Rebel. I can't do anything with single versions anymore. That was yesterday and it was good then. Not today.

[RubiChess]

I expect programmers to adapt to reality as well.

There are thousands of engines out there. If one doesn't fulfill your expectations, just leave it.

This kind of attitude rose in the last year in this community. The engines are for free, and the authors can do what they want. And I bet there would be much more nice engines, if programmers wouldn't have to read this kind of stuff all the time. I know several programmers who quitted because of this.

It's just Nörgel-Ede (my private nick for him) and no programmer should worry too much about this stuff.

Regards, Andreas

[Eduard]

There are such positions where little power logically also plays badly:

Relax_,S - Solista,SugaR AI 2.50 1-0, Rated game, 5 min 2022
[fen]r3q3/1b1nrpk1/2pb2p1/p2p3p/Pp1P3P/1N2PNP1/1PR2P2/2RQ1BK1 b - - 0 1[/fen]


Analysis by SugaR AI 2.50 sse41-popcnt (on two old cores Laptop):

40...Rxe3 41.Nc5 Bxc5 42.fxe3 Bd6 43.Qe1 Nf6 44.Bh3 Ng4
-+ (-1.62) Depth: 10/13 00:00:00
40...Rxe3 41.fxe3 Qxe3+ 42.Kg2 Qxb3 43.Bd3 Re8 44.Re2 Rxe2+ 45.Qxe2 Nf6 46.Qc2 Qxc2+ 47.Rxc2 Ne4
-/+ (-0.98) Depth: 11/22 00:00:00
(---)
40...Rxe3 41.fxe3 Qxe3+ 42.Kg2 Qxb3 43.Qd3 Qxd3 44.Bxd3 Nf6 45.Rxc6 Bxc6 46.Rxc6 Rd8 47.Ra6 Ng4 48.Rxa5 Ne3+ 49.Kh2 Nd1 50.Bc2 Nxb2 51.Bb3 Rc8 52.Rxd5 Rc3 53.Rxd6 Rxb3 54.Ne5 Nxa4 55.Rd7 Rb2+ 56.Kh3 Rf2 57.Rb7 Nc3 58.Rxb4 Ne2 59.g4
+/= (0.43 --) Depth: 30/42 00:02:01 101MN, tb=100

My engine played here with 2 cores on an old Intel processor! Only 850 kns with Stockfish 14. On a modern PC that is 20x faster, the mistake Rxe3? does not happen. Is the engine bad because it made a mistake here? No. The PC is no longer up to date. Testing engines on old processors or on just one core and then allowing oneself to conclude whether the engine is good or bad is not my thing! Feel free to continue with your 1core test games, I'm just doing this for fun so my old laptop PC doesn't rust. There is no more sense!