A while back there was a discussion about the frequency of indexing
errors, and how they were usually a byproduct of the development
cycle. Well, I have a reproducible error, now, that isn't.
Here's what I do:
1. Import references into my database (about 10 MB).
2. The db opens and works just fine.
3. Close the db window.
4. Reopen the window, it works fine.
5. Now, do a SQL search. This forces indexing to occur (the first
search takes about a minute, after that seconds).
6. Close the window.
7. Reopen the window -> error 372!
I've done this three times, on OS 9. It's a single file, db, BTW.
I think that the indexing is probably going well, but that an error
is made when writing it to disk. (I've alloted 27 MB for RB with
this, BTW).
Ruslan, I think you can reproduce this with the project I sent you,
if I send you the file to import. Would you like me to do that? It's
about 1.2 MB stuffed (+ binhex overhead). I can put it up on an ftp
site, if that would be better.
Jon
©2001 jda |
