With either CompareIt Unicode 3.8.1.1652 or CompareIt 8-bit 3.8.1633, with no "Ignore" options set, compare the a.db to b.db from http://home.arcor.de/dandv/bugs/compare ... nclean.zip and merge the "virtual lines" at the top of a.db over the text in b.db.
Open b.db in any hex editor (Necromancer's Disk Navigator would be great) and notice that, binary speaking, b.db was messed up:
- most 0x00 bytes have been replaced with 0xB7
- the rest of the 0x00 bytes have been replaced with 0x0A !
Even worse, if you compare a.db and b.db again, both CompareIt!s will say the files are identical!
Hope that helps,
Dan Dascalescu
--
Check out my funny Californian pics at
http://whatsnew.fotki.com/harwons
(bookmark this address and stay tuned to my latest pics :-)
Fatal bug: CompareIt! not binary clean: 0x00's replaced
I have custom-format database files and records are separated by 0x00 bytes, and rows are separated by 0x00, 0x00, 0x0D, 0x0A.grigsoft wrote:Thank you for your report!
In fact 0 chars are not allowed in text files
You can find a sample at http://home.arcor.de/dandv/bugs/compareit/sample_db.zip
When you compare the two files, you will see why I would need at least patching, if not merging