>Well, actual re-compile is certainly not an issue, except that in our environment, where building includes updating Source Control and what not, and installing it for our users, it is much more of a hassle than anybody would like.

I know, which is why I have the whole process automated, including the source control. The only downside is that I may forget I launched the process (when I added checkouts and checkins, it went from 1:30 to whole seven minutes), and that we may be doing too many builds now that it's automated.

>Beyond that, this happens (has happened) in quite a few different applications, so re-compiling all of them actually IS quite a bit of work.

I meant to suggest to do this on the two or three where the error is most frequent.

>So your guess (and everybody else's, it seems) is that the problem is with datatoclip somehow -- that occasionally, operating on exactly the same data, it fails to produce anything?

Actually, no, I wanted to get it off the list of suspects. Or if it was misbehaving in a way, this would be the way to catch it - hopefully the contents of the clipboard at the moment would give you some clue as to what's going on.

My guess is with Al's, that Excel is unreliable, and I'm glad he procured that link before I did, otherwise I would. I found that article while fighting the .doc-to-.pdf problem, where Word would behave worse and worse in every new version, and I had eventually bought TextControl (which can load .doc and .docx files with minimal memory consumption) - and they have that article linked on their main page. The main problems with Word were that it would eat more memory with each new document loaded, adding about 20M (and the files were 1-2 pages each), returning none of it when document was closed. So I weighed the speed vs risk of crash and decided to kill Word on every 20 documents and instance a new one. Still not good enough - it would complain if something was wrong in the merged document (and these all were merges), if there was something in the unprintable area, and the dialogs weren't suppressable, as they weren't on the list, or of the same class (if it has classes at all) as those we could suppress before.

This is just to illustrate what Word can do to your process. With Excel, no such recent experience - luck! - but I had a friend ask me for advice, and it involved exactly the _vfp.dataToClip(,,3) trick, which he'd paste into a named range. The named range has two rows, and before the paste, the top left cell of the range is selected, and .addRow() calls are issued recno() of times to heighten the range to accept all the rows. Now these calls just don't add rows (not sure I understood what he was doing, looks somehow different from what I told him, and I don't have excel at home to try - in 2004 this worked flawlessly). I've seen other things which just silently stop working from one version to the next.

I think the jury can't stay out this long, it's getting rather cold.