True or false? Chess will never be solved! why?

Those questioning what it actually means to solve chess, or that we must first agree upon definitions actually seem to have a pretty salient point:

There seems to be a lot of not understanding what "solving chess" actually means.  It's not up for intepretation.

Semantics shouldn't come into it.  The mathematical definition of what it means to solve a game is not up for debate.

This kind of vocabularily undermines anything that could be considered a proof.  No, I'm sorry, guiding principles need not apply; this exercise is about proof, not approximation.  You can't discard an entire subtree because it looks like it's pretty much determined -- you must prove it, or the rest of the proof crumbles around it.

This 'vocabularily' is required when one is talking about hard values that have not been calculated yet.  Believe me, I am not talking about discarding parts of the tree based on guiding principles.  Also, I should mention that everyone assumes that traversing this vast tree will have to be done from the starting point, and thus the tree expands outward endlessly as each move accumulates.  It wont be.  That's how computers do it now because they are trying to play Chess, not solve Chess.  

It will be traversed backwards from the very finite set of mates that can exist...even more finite when you factor in knowns from the starting position.

False, Moor's Law, refuted to a draw in the Ruy Lopez, Scotch, King's Gambit (though white will struggle a bit for it), Queen's Gambit, Reti, Queen and King's Indian, Leningrad Dutch, Stonewall Dutch, various anti-Dutches and the Manhattan Gambit, Caro-Kahn, center counter, Benko Gambit, Benoni Defense, Nimzo-Larsen Attack, English, Grob, and Alekhine's Defense.

Refuted to a win for white against the Latvian, Borg,and St.George's Defense.

Win for black against the wing gambit, From's Gambit, and Danish. 

Up to a point.  You are correct that solving chess is a theoretically acheivable exercise.

It's the practicality that is, and will forever be, out of reach.

Indeed that is exactly what's taking place with the generation of tablabases -- solved compeltely up to seven peices now.

In fact, a solution would effectively be a 32 peice tablebase.  The problem is that it will never (and can't) be calculated.  The game tree complexity puts it entirely out of reach.

I don't think you truly appreciate the magnitude of the problem.

See above.

I don't think TCJ understands anything, really.

Similar to time travel? 

I'm travelling through time right now.

Sure, but try traveling the other way TheGrobe!

Memory.

True. Since even if the computer may have already solved it, humans will never find out anyways. Humans will never fully absorb chess. Besides, speaking of computers, for humans chess is not only technical. 

Indeed that is exactly what's taking place with the generation of tablabases -- solved compeltely up to seven peices now.

In fact, a solution would effectively be a 32 peice tablebase.  The problem is that it will never (and can't) be calculated.  The game tree complexity puts it entirely out of reach.

Entirely out of reach for the brute calculation of a 32 piece tablebase, minus any culling at all, and given the current silicon processor-based extrapolation of the future of processing power.  That much I will agree with.

What do you mean wrong?  He was right, "The Turk" was a phoney.

"You said something isn't achievable, that's the same thing as saying everything isn't achievable!  Where would we be without progress?  Nowhere!  That's where."

"You're wrong if we change the accepted definitions!  Like what 'solved' means.  Or 'proof'..."

"Someone else was wrong about something similar once, so you must be wrong too."

The mind reels...

I think a lot of people here talking about how we can slim down the tablebase don’t understand how numbers work.

There are estimated to be about 10^43 legal board positions. Let’s say that we came up with a really awesome way to cut out positions where the result is obvious, and this method allows us to eliminate 99.9% of the positions from analysis. So now we only have to analyze 0.1% of the possible positions. How many positions does that leave us to still analyze?

10,000,000,000,000,000,000,000,000,000,000,000,000,000

Good luck with that.

Yeah, when dealing in such enormous orders of magnitude, cutting half, or even three quarters of the work doesn't really help (often it won't even change the exponent!).  Cutting 90% gets you from 10^123 to 10^122.  Cutting 99% gets you to 10^121 and 99.9% to 10^120, none of which really make the problem any more manageable.

I don't get it?  If you write it enough times it becomes fact?

Well, I can agree with you about the tic-tac-toe part, anyway.