anything else that can be done here without trial & error? 
xjames KwonTom Obsessive Puzzles: 1974 Best Total: 31m 41s  Posted  2008.03.19 03:39:11 Hi... I almost finished this puzzle without any guessing or trial and error, but hit a snag here. I already finished after trial and error, but I was wondering if there was anything else to be done here that I missed.
Thanks for any thoughts.
Oh, I saw another nontrial and error thing that I added after I captured the screen.

prj KwonTom Obsessive Puzzles: 2356 Best Total: 18m 20s  Posted  2008.03.19 04:05:45 The 1 at r6c16 must have a line coming in from the top left. Since there's a 2 there, it must also have a line coming in from its top left. You can get 4 lines from that, and a couple more to prevent the loop from closing.
Last edited by prj  2008.03.19 04:06:02 
Naivoj KwonTom Addict Puzzles: 314 Best Total: 33m 50s  Posted  2008.03.19 07:34:59 Everything is forced using color and path deductions, although my last path deduction is not that easy: 20060520SatNaivojColorTips
Last edited by Naivoj  2008.03.19 07:53:58 
xjames KwonTom Obsessive Puzzles: 1974 Best Total: 31m 41s  Posted  2008.03.19 11:07:54 hey guys, thanks for the thoughtful replies. I do get Prj's lines, although it took me a bit of looking to figure it out.
Naivoj, I got most of your recommendations, but I got lost with the whole colors thing. And I really appreciate the time you took to make that word doc. I studied it with care, but the whole colors bit was lost on me.
I'll look at it again tomorrow, see if a fresh mind helps... the whole thing's starting to look fuzzy to me at this late hour... lol
thanks again! 
PuzzleScot KwonTom Obsessive Puzzles: 2653 Best Total: 19m 5s  Posted  2008.03.19 13:00:05 I'll have a go at explaining 'the colours thing'  I use colours all the time (How do I add in my message grid below?). Red for 'inside' the loop, and yellow for outside. [At the edge of the grid, there is effectively a ring of yellow grid squares, for the sake of the cases below.]
Consider the colouring of the 4 squares horizontally and vertically adjacent to our square: Around a '3', There is 1 square matching, and 3 opposite around a '2', 2 matching and 2 opposite around a '1', 3 matching and 1 opposite around a '0', 4 matching, as we all know... (Stating the bleedin' obvious, I know, but it helps in the next bit..)eg,
If the 2 '?' squares at r3c1 and r4c2 are say both RED, then we know r2c2 and r3c3 are both YELLOW. Also, r1c3 and r2c4 are both RED.
Also, if we have r3c1/r4c2 opposite colours to each other, then we also know that r2c2/r3c3 are opposite colours, as are r1c3/r2c4
A typical example is a start grid with a 2 in the corner. It already has 2 yellow neighbours, so the remaining 2 neighbours must be red.
Similar idea with '1's  once 2 of its neighbours are of the same colour, the '1' must be that colour too. If we colour 2 neighbours opposite colours, the remaining 2 borders must be 'x'...
Hope that helps... Alan 
xjames KwonTom Obsessive Puzzles: 1974 Best Total: 31m 41s  Posted  2008.03.19 22:38:41 actually it does help..
I'll try doing some archives with colors to try and get the hang of it.
Thanks 
Naivoj KwonTom Addict Puzzles: 314 Best Total: 33m 50s  Posted  2008.03.20 01:14:08 The most basic color rules are:  2 adjacent squares separated by a Cross must be of the same color  2 adjacent squares separated by a Line must be of opposite colors
There is a harder color pattern, I did not explained separately as a pattern in the word document, regarding the line deduction to the right of the clue 3 by the corner:
Clue 1 and 3 Same Color Pattern  Requirement: the two ? squares are of the same color.  Results= 1 Line, 2 Crosses and 1 Color
The main result is the 1 Line. Once you get that Line the other results can be deduced from simpler rule/pattern, like the "Clue 1 Opposite and Non Adjacent Color Pattern" I explained in the word document.
If the two ? squares are of opposite colors it becomes the Clue 1 and 3 Opposite Colors Pattern which gives 2 Lines, 1 Cross and 1 Color 