Andscacs

[cdani]

Hello!
Yes, unless I had magically access to hundreds of cpus, I don't plan to continue with it. Or maybe in some years with much more powerful computers, who knows...
Many thanks to all for these nice years!
Best regards.

[Guenther]

Hello!
Yes, unless I had magically access to hundreds of cpus, I don't plan to continue with it. Or maybe in some years with much more powerful computers, who knows...
Many thanks to all for these nice years!
Best regards.

Well, what a pity. Thanks for all your work and participation, Daniel!
I hope you will come back in a few years.

Best regards,
Guenther

[Alayan]

Yes, unless I had magically access to hundreds of cpus, I don't plan to continue with it. Or maybe in some years with much more powerful computers, who knows...

CPU hardware for idea testing is indeed quite limiting.

Noobpwnftw is donating hardware time to Fishtest and OpenBench, and some engine authors (Demolito, RubiChess, Weiss) have joined OpenBench. It makes much easier for them to not be bottlenecked by testing time, especially as authors usually don't all test things at the same time. Maybe you could think about it.

Speeding up the engine is also useful, as it gives you a higher effective TC at equal resources.

That being said, trying to gain elo for a strong engine is quite unforgiving. Just doing the same stuff as everybody else is not so interesting, the thrill is to discover new concepts that gain elo and hopefully could also help to move the field as a whole. Computer Chess is waiting for the next search trick...

[Ovyron]

Thanks a lot for Andscacs, it was the strongest engine with original style that I tested.

Also, guys, it's open source! If you don't want Andscacs's flame to extinguish there's nothing stopping YOU from releasing next version

[Alayan]

Incorrect, the last open source version was 0.921 ; while the latest released version was 0.95 (~20-25 elo stronger).

I downloaded the 0.921 sources a few days before andscacs.com went down ; I could upload them if other people are interested and don't have them.

[Ovyron]

Incorrect, the last open source version was 0.921 ; while the latest released version was 0.95 (~20-25 elo stronger).

Well, here's the plan:

1. Add some ~20-25 elo so it goes up to 0.95's strength.

2. Pretend this new engine was the last one before Daniel quit.

3. Add another ~20-25 elo and call it Andscacs 1.0.

4. The torch lives on!

Seriously though, the community has done amazing things with open source software, I'd be much more sad if this was closed source and doomed (like so many other chess projects that people abandoned without opening the source).

[Alayan]

If someone wanted to make live on a dead open source engine, it would be easier to take Laser and build on it. Source in english and stronger.

[Ovyron]

I didn't know Laser was abandoned. I never had time to check it at all. But yes, in general I'd like to see a revival of old dead projects, for old time's sake. Open sourcing is a great tool and it'd be a shame if most open source engines were never visited again by someone in the future and being open was useless.

[Roland Chastain]

Incorrect, the last open source version was 0.921 ; while the latest released version was 0.95 (~20-25 elo stronger).

I downloaded the 0.921 sources a few days before andscacs.com went down ; I could upload them if other people are interested and don't have them.

I would be interested.

[jpqy]

You can check here and work on these engines when interested :

http://chess.grantnet.us/

Would be nice if more open sources engines could be included in this testframe.