Re: Btrieve problems. Please help!!!!

Kevin Shank <kevin@gcs_npd1.dscinet.com> wrote: >I cannot help you with this specific problem, however I was hoping you may be >able to help me since you sound like you know your way around btrieve. Our >customers who are using the software which I developed are reporting an >alarming number of database corruption problems. They are getting I/O errors >which cause some of the records to be unaccessable until the database is >recovered and loaded back into a clean database. The platfrom is WFW3.11 and >the problem has been reported by folks using 5.10 and 6.15 btrieve. Any >suggestions? Thanks a million. >E-mail to kshank@galaxy.dscinet.com OS/2 at one time had a -- well, not bug, but condition -- that allowed chkdsk to corrupt btrieve files; it was related to btrieve's pre-allocation of file space. There was a chkdsk fix available. Thought it had been fixed in Warp3, but not sure about WFW. Will check into it and let you know something by day after tomorrow. Quick temporary fix on that one was to remove the partition with the data files from the "IFS=...autocheck:..." line We also ran into some wierdness under OS/2 & btrieve 5.1 if the number of pages in a file grew to 32767; apparently 5.1 is still using a 16-bit page pointer. Our fix was to rebuild the offending files with a larger page size. Hope this helps, -- Rand McNatt (rand@netjava.com)