Posted: Wed Aug 17, 2005 3:20 pm Post subject: Really want to give up
I'm new here. I'm also trying to develop a good sudoku solving program. I was doing all the work on my own, and didn't know anything about you guy's progress. I thought I have finally create a really powerful program, so that's why I'm here.
But, the only thoughts I've come up after reading the threads in this forum is to give up my research. Why? Because even I have done some much effort on it, I found that I still haven't invent any new ideas differ from those ideas you guys are discussing here.
For example, my so-called "tree analysis" is only a stricted form of the "Trebor's Tables" you guys have here; the only different is, I never check if a case tree includes a contradiction ( because I consider this to be trail-and-error ) , but only take the intersection of all the case trees. Indeed, tree analysis is a extremely powerful tool, and so far no puzzles remains unsolved using this method. But anyway, you guys have already known that, right?
I was quite satisfied by this method, but obviously some of you still not. You guys seems to have a greater ambition than me, which makes me feel powerless. I can't see any reasons why should I keep working on this, since we have so many better developer here.
So... Well... I guess that's all guys, and thanks for your attention.
I thought much the same, but I still enjoy working out the logic behind each process and coming up with my own implementation. I'm working on the theory that as I understand more facets of each solving technique I'll eventually get a good enough grasp to suggest improvements, or spot better/different ways of doing things.
One day, with enough code in my solver, I'll have some good ideas to contribute.
I'm not even trying to figure out a good logic. Yet I think I've done atleast something. Thus the primary question is: "Why are you doing sudoku stuff in the first place?" - you've already done something remarkable not everyone are able to do. It is different than the other methods and you made it. Heck, all I have in my own solver is three of the simplest possible logics out there and a backtracker. I made it all fast because speed was a concern. I combined it to be something original. That is what matters, you did something and it is original. The better you did it without help. I spend a good while reading up on all the logics they have here, but that was pretty much in vain. I didn't have the capability to do the logics even by reading about them.
So, you are one of the people who have done a logic and are able to discuss well about them. Afaik nobody here is in advantage. Your method of doing it is equal to others.
Thanks; I think I'm feeling much better now.
My original goal is to find a set of strategy not involving trial-and-error which can solve all valid sudoku puzzles, and after I developed the tree analysis, I supposed that I've done. But, I guess some of you won't agree with me that tree analysis isn't a trial-and-error method. That's right; tree analysis does not involve "error", but still it involves "trial". Although under my definition for the trial-and-error method it is not, but I know some of you might consider a real non-trial-and-error method to be not even involving any case-tree-generation. Though personal I won't completely agree with that, but my main goal is to construct a persuasive strategy for everybody, not for me only. I'm sure what's you guys' opinion for this.
hehe, you don't have to be the best programmer in the world !
Maybe it's also quite good to be the best in your village ;-)
And when you start a project you can't know in advance whether
it will develope something good or new, but of course you
should try nevertheless.
When you started a promising task - which then later turns
out to be useless _by accident_ , meaning that you had no
chance to influence it (e.g. because someone else has a
faster program) then that shouldn't depress you, because
you did the best and made no mistake.
That's much better than winning the contest with boguous code
just being lucky that the samples which your solver couldn't handle
were missing ! (IMHO)
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum