no more ChessGUI

Discussion of anything and everything relating to chess playing software and machines.

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
User avatar
Evert
Posts: 2923
Joined: Fri Jan 21, 2011 11:42 pm
Location: NL
Contact:

Re: no more ChessGUI

Post by Evert » Fri Apr 01, 2016 7:20 pm

Alexander Schmidt wrote: FRC is one of the most played chess variants. Don't you think without adapting these present-day reality would Fairy-Max (the variant engine) keep stucked in the stone age? :)
If Fairy-Max would play FRC and someone's only interest in Fairy-Max would be that it plays FRC, they are rather missing the point, in my opinion.
It's capable of playing dozens of variants, many of which differ from FIDE chess in more interesting ways than shuffling the back rank pieces and adopting an obscure castling rule.

Alexander Schmidt
Posts: 1086
Joined: Thu May 10, 2007 12:49 pm

Re: no more ChessGUI

Post by Alexander Schmidt » Sat Apr 02, 2016 8:47 pm

hgm wrote:
Alexander Schmidt wrote:FRC is one of the most played chess variants.
And your claim is perilously close to an April-fool joke. From the Chess servers I know the only one where FRC is the most played variant after orthodox Chess is a side where it is also the only variant they offer besides orthodox Chess.
One of the most played variants...

Alexander Schmidt
Posts: 1086
Joined: Thu May 10, 2007 12:49 pm

Re: no more ChessGUI

Post by Alexander Schmidt » Sat Apr 02, 2016 8:51 pm

Evert wrote:
Alexander Schmidt wrote: FRC is one of the most played chess variants. Don't you think without adapting these present-day reality would Fairy-Max (the variant engine) keep stucked in the stone age? :)
If Fairy-Max would play FRC and someone's only interest in Fairy-Max would be that it plays FRC, they are rather missing the point, in my opinion.
It's capable of playing dozens of variants, many of which differ from FIDE chess in more interesting ways than shuffling the back rank pieces and adopting an obscure castling rule.
I tried to explain HGM with this example that programmers of freeware programs are free to not implement something. HGM complains again and again that Arena misses a feature he would like.

Nevermind :)

User avatar
hgm
Posts: 23619
Joined: Fri Mar 10, 2006 9:06 am
Location: Amsterdam
Full name: H G Muller
Contact:

Re: no more ChessGUI

Post by hgm » Sun Apr 03, 2016 6:38 am

This is because you still miss the point, as your example shows. Whether Fairy-Max or any other engine supports a particular Chess variant has no effect on anything else. Not supporting it hurts no one, and in particular does not hurt people that do not use Fairy-Max at all.

On the other hand, when popular GUIs adopts a standard that is not inter-operable with the rest of the world, this will cause engines that use this deviant standard to be created. Meaning that people using other GUIs now also cannot use these engines. So people that do not use the offending GUI, and have no interest in using it at all, are still hurt by this through not being able to use the engines they want. And it is especially bad when this deviant standard is defective. Then it doesn't only promote chaos, but also misery. If your actions have a large impact on what others do, there is a moral obligation on everyone to not refrain from actions that on average are damaging and hurtful.

A good analogy would be where GUIs are restaurants, protocols are credit-card companies, and Chess varians are food dishes. The existence of different brands of credit cards becomes a great annoyance to customers, who now have to walk around with an enormous wallet full of cards to prevent the get to hear "sorry, you cannot eat here because we don't accept your brand of credit card". Of course people could try to blame the restaurants for this, assuming that every restaurants should accept every brand of cresdit card, loading the burdon on the entrepreneurs.

But now suppose that credit-card company X is known to swindle its users, charging them 50% more than they actually pay. Then I would not think it would be great if customers could also pay with X there. In fact such restaurants would set up themselves as acccomplishes for duping the customers, and give them the impression that card X is as good a means of payment as any other.

Your Fairy-Max example in this analogy would be that an ice-cream parlor (Fairy-Max) should better also serve rice, because rice is the most consumed food in the world. Are you really surprised that such an example does absolutely nothing to revise my optinion on credit card X as thieves and crooks? That I would think it is OK for them to steal from their customers because I do not serve rice???

Alexander Schmidt
Posts: 1086
Joined: Thu May 10, 2007 12:49 pm

Re: no more ChessGUI

Post by Alexander Schmidt » Sun Apr 03, 2016 2:34 pm

hgm wrote:chaos...misery...damaging...hurtful.
The implementation of Chess960 in Arena does what it is supposed to do: It is possible to play Chess960. Your assertion Arena brings chaos and misery over the world because there is a problem in very rare positions while analysing Chess960 fen positions without knowing the previous moves is simply burlesque. :lol:

User avatar
hgm
Posts: 23619
Joined: Fri Mar 10, 2006 9:06 am
Location: Amsterdam
Full name: H G Muller
Contact:

Re: no more ChessGUI

Post by hgm » Sun Apr 03, 2016 2:47 pm

The misery, however, is that engines written according to Arena standard will NOT be able to play Chess960 on most other GUI. Apparently an issue important enough for several people including yourself to call ChessGUI 'great' for being able to run them, while GUIs that cannot run such engines are merely 'fine'. So the total amount of hurt/damage by your own words is that it reduces the user experience for almost every GUI that is in common use from great to fine.

I would say that is significant damage. And for what? For the right to pester Arena users with the problem that most compliant UCI engines only work using Polyglot as an adapter, and that in some (fortunately rare) cases the position will be misunderstood. The latter is indeed not the main dish, but just the 'icing on the cake'.

That doesn't sound so burlesque to me...

Modern Times
Posts: 2387
Joined: Thu Jun 07, 2012 9:02 pm

Re: no more ChessGUI

Post by Modern Times » Sun Apr 03, 2016 2:57 pm

Alexander Schmidt wrote:
hgm wrote:chaos...misery...damaging...hurtful.
The implementation of Chess960 in Arena does what it is supposed to do: It is possible to play Chess960. Your assertion Arena brings chaos and misery over the world because there is a problem in very rare positions while analysing Chess960 fen positions without knowing the previous moves is simply burlesque. :lol:
The more standards a GUI can support the better for the users, and they are the ones who matter. So I applaud ChessGUI for supporting the Arena implementation as well as the others. HG may not like the Arena implementation, but it exists and it is a popular GUI.
.

Opinions expressed here are my own, and not necessarily those of the CCRL Group.

kbhearn
Posts: 411
Joined: Thu Dec 30, 2010 3:48 am

Re: no more ChessGUI

Post by kbhearn » Sun Apr 03, 2016 3:20 pm

hgm wrote:The misery, however, is that engines written according to Arena standard will NOT be able to play Chess960 on most other GUI. Apparently an issue important enough for several people including yourself to call ChessGUI 'great' for being able to run them, while GUIs that cannot run such engines are merely 'fine'. So the total amount of hurt/damage by your own words is that it reduces the user experience for almost every GUI that is in common use from great to fine.

I would say that is significant damage. And for what? For the right to pester Arena users with the problem that most compliant UCI engines only work using Polyglot as an adapter, and that in some (fortunately rare) cases the position will be misunderstood. The latter is indeed not the main dish, but just the 'icing on the cake'.

That doesn't sound so burlesque to me...
Technically UCI only needs FEN to be able to express initial positions, since it follows that up with a movelist. so as a protocol to play 960 it's not defective, just irritating that you can't specify an arbitrary midgame position as an arena960-fen.

User avatar
hgm
Posts: 23619
Joined: Fri Mar 10, 2006 9:06 am
Location: Amsterdam
Full name: H G Muller
Contact:

Re: no more ChessGUI

Post by hgm » Sun Apr 03, 2016 3:36 pm

Virtually every GUI allows people to set up an arbitrary position on the board. If a protocol breaks that function, forcing people to set up the position by a sequence of legal moves from a start position instead, I would say it does a lot of damage. UCI is not just a protocol to play complete games. It was supposed to be a interface definition for ordering engines to do what people want typically done. One of those things is setting up arbitrary positions from pasting FENs (e.g. mate problems).

Alexander Schmidt
Posts: 1086
Joined: Thu May 10, 2007 12:49 pm

Re: no more ChessGUI

Post by Alexander Schmidt » Sun Apr 03, 2016 3:41 pm

hgm wrote:The misery, however, is that engines written according to Arena standard will NOT be able to play Chess960 on most other GUI.
Yes, most likely because most of the engines where written before there where other specifications. Just to remind you, speaking about Winboard engines: The first attempt to run Chess960 engines was done by Tim Mann in Winboard. Arena simply did it the same way as it was done in Winboard.

Post Reply