World: r3wp

If that were the case, the 'layout would fail ...

When are the paths in the layout evaluated? The only part I don't 
know is the View part.

Immediately

In that error there are only 3 path roots that could be none: db, 
db/:counter and FN. If the db is empty then db/:counter would be 
invalid even in the layout. That leaves FN.

This seems like a job for PROBE.

If ''layout completes, then FN will be a VID field object

so this suggests to me that db = [ ]

If there are no rows returned then that makes sense. Looks like the 
database access needs some work. And even if the path succeeded, 
the SQL at the end of btn-save would fail because the database is 
disconnected. Plus, the create statement at the beginning just creates 
a database, not a table.

I didn't look at the rest of the code .. I'm just interested in the 
tech explanation as to why the function errors out like that

The error message itself refers to some fairly limited code. A couple 
probe statements would narrow things down even further.

Anyone know which algo is used by the SORT function ?

Not a stable one (there's a ticket about that). It could be guessed 
by using a logging compare function.

So, did you guess it ?

I would bet on a quick sort.

If it uses several algos it will be difficult

Quick sort is not optimal for cases where the comparison cost is 
expensive.
And with Rebol comparisons have high cost

>> print mold collect [sort/compare [20 19 18 17 16 15 14 13 12 11 
10 9 8 7 6 5 4 3 2 1] func [x y] [keep/only reduce [x y] x < y]]

[[10 20] [1 10] [10 20] [10 19] [10 2] [10 18] [10 3] [10 17] [10 
4] [10 16] [10 5] [10 15] [10 6] [10 14] [10 7] [10 13] [10 8] [10 
12] [10 9] [10 11] [10 11] [10 9] [5 1] [9 1] [9 5] [5 2] [5 3] [5 
4] [5 6] [5 8] [5 7] [5 6] [5 4] [6 7] [7 8] [8 9] [6 7] [7 8] [6 
7] [1 2] [2 3] [3 4] [1 2] [2 3] [1 2] [16 11] [20 11] [20 16] [16 
12] [16 13] [16 14] [16 15] [16 17] [16 19] [16 18] [16 17] [16 15] 
[17 18] [18 19] [19 20] [17 18] [18 19] [17 18] [11 12] [12 13] [13 
14] [14 15] [11 12] [12 13] [13 14] [11 12] [12 13] [11 12]]

ok it's quick sort

far from behing the best one in our context as I said. Too much comparisons

*from beeing

Darn. Quicksort is still slow for mostly sorted stuff:

>> print mold collect [sort/compare [1 2 3 4 5 6 7 8 9 10] func [x 
y] [keep/only reduce [x y] x < y]]

[[6 1] [10 1] [10 6] [6 2] [6 3] [6 4] [6 5] [6 7] [6 9] [6 8] [6 
7] [6 5] [7 8] [8 9] [9 10] [7 8] [8 9] [7 8] [1 2] [2 3] [3 4] [4 
5] [1 2] [2 3] [3 4] [1 2] [2 3] [1 2]]

See also this: http://curecode.org/rebol3/ticket.rsp?id=1152

Merge sort seems to give the best results when the comparison functions 
are expensive.
http://warp.povusers.org/SortComparison

Good idea, as long as the sort is stable.

Shell sort seems pretty good aswell

Merge sort is for when comparisons are *really* slow, sorting disk 
records slow. You have to have a really heavyweight mezz compare 
function to justify that. The overhead of the second series would 
be pretty high in cases where the speed difference would really matter. 
One of the inplace sorts would be better, as long as it's stable.

there's no need to have heavy mezz, the Rebol's scheme is already 
50 times slower (at least) than same compiled code when it ferdorms 
comparisons.

std::sort seems to do really well.

I don't know this algo

Apparently the actual algo varies from compiler to compiler, but 
it is a mix of qsort, heapsort and insertion sort depending on which 
is fastest in particular circumstances.

But it is in the C++ standard library, so C code wouldn't be able 
to use it without copying and rewriting.

It could in theory be the best, but that depends on how smartly it 
was written. It tends to be pretty reliable though because the pathalogical 
cases are in theory sorted with different algorithms when they would 
otherwise cause problems.

std::sort is not stable.

std::stable_sort is.

Interesting. It's been a long time since I used C++ (there were no 
standard libraries then). It never occured to me that someone would 
use an unstable sort algorithm without checking first whether it 
would be safe to do so. I must have missed that in college.

the funny thing about REBOL SORT is, that, it is actually stable, 
provided you use /compare

, which proves, that it just uses a wrong operator somewhere

I don't think it's the purpose of compare

That is what I thought when I write the ticket. Could you give an 
example of /compare use which is stable, to add to the ticket?

/compare is used to sort your own data structures

Or to reverse a sort, or for whatever you need it to be for.

OK, Brian, will do

For reversing we'd also have /reverse.

Yeah, we have options for the most common variants, and the default 
for the most common case. The /compare option is for everything else.

And not well enough documented either. I don't really know how /compare 
works, not completely.

to sort  gobs for instance, with your own scheme

I know *why* it is there, I just don't know how it *works*. How does 
it react to different return types of the comparison function? What 
types are allowed? When is the security hole going to be fixed?

I know, that it is documented somewhere, that the Comparator can 
yield either logic! value or a number, where a negative number means 
"<", zero means "=", and a positive number means ">"

Btw, Linux A110's SORT seems rather stable for me :)