Danny,
If you're responding to my posts, and I'm responding to yours, I think it's a dialog.
It seems you are designing a standard to fit your parser rather than designing a parser to fit your standard. I can't even remember how many parsers I've designed and implemented over my lifetime. At least 4-5 were heavy duty. Probably another 30-40 were light duty, like the one we are discussing. I prefer language and notation whose objectives are a good compromise with the needs of a human interface. In this case, we are dealing with columns, which are essentially fixed position. Space delimited tokens ALLOW you to line up the columns, but they don't guarantee it. From a human interface point of view, it seems like a kludge.
I also prefer a mid-way layout that can be played both backward and forward.
As for the mind-boggling complexity implied by embedding several layouts of the same inital game or layout, along with partial solutions, I see no added value over a series of independent (context free) layouts that are terse and visually clear to a human.
I'll use game #6240 and your example of Shlomi's solution, along with my current solver's solution, as examples of my proposal in my next post.
-Gary
From: dannyjones183
Sent: Sunday, December 16, 2012 5:39 PM
To: fc-solve-discuss_at_yahoogroups.com
Subject: Re: Monologue: FreeCell Exchange Format
--- In fc-solve-discuss_at_yahoogroups.com, "Gary Campbell" <gary_at_...> wrote:
>
> Just 4 shoot-from-the-hip points in response (so, I guess itâs no longer a monologue).
Correct. It's not even a dialog where we're talking independently about the same topic.
> 1. Do we really need âbeginâ ... âendâ keywords? Why not just begin with the Game# and freecells: lines?
Yes, we really need the Begin/End keywords. The reason is obvious. We are discussing a FreeCell solution, and there are blocks of information that belong together as a unit. The transition from one block of information to another is distinctly identified. Identifying it as a FreeCell block forces the exchange format for the remaining data.
> 2. Are ** characters really better than space-space characters? Why?
Yes, the ** characters are really better than your space-space characters. The reason is because the input is space-delimited tokens that can be parsed. Space-space only works with fixed-position formatting. I inserted extra spaces in Moves to make it easier for people to read. But, those extra spaces aren't necessary in the exchange format.
> 3. Have you ever run into a case where other than 1 or the max number of cards needs to be moved to an empty column?
I've never generated such a move because my solver only generates moves that are acceptable to FcPro. However, that doesn't mean that someone else might generate such a move. I just don't have a usable example handy. Shlomi's example in game #31465 appears to be extraneous because it's a maximal move as well.
> 4. Also, Iâm curious: Looking at the 2nd âtableauâ the cards topping columns 2-5 are the same as the first tableau, so I assume the 2nd example is a continuation of Game #6240. If Iâm correct, I have trouble seeing how the solutions connect to one another. Or, why the remainder of the 2nd solution is actually longer than the entire solution of the first example? It seems that a notation should allow for a mid-way layout and the steps from some earlier layout that got you there.
My example shows the progression of Shlomi's original solution. There's only one game being presented between the #Begin FreeCell and the #End FreeCell. Also, the initial contents of the table can be presented with either a #Game record or a #Begin Layout/#End Layout block. I showed both for completeness. I then listed the first 72 moves of Shlomi's solution. This brought us to the second layout showing the state of the table. I then presented the remaining moves in his solution. I chose this example because I had all of this information from earlier discussions about his solution and Horne automoves when he makes his 73rd move.
>
> Would anybody like an example of my proposal in the same form as Dannyâs, or are the differences âvisibleâ?
>
Yes, an example would be nice. I had to sit down and "map out" your proposal when I read it.
Received on Mon Dec 17 2012 - 07:43:00 IST