On Monday, December 30, 2002, at 10:14 AM, Brian Abbott wrote:
> I don't think the fpt corruption is to do with Codemine specifically.
> This sort of thing is likely to happen with any vfp (indeed to a
> greater
> or lesser extent any database) application if you have the sort of
> power
> problems you describe. Solution is simple: tell them they _must_ spend
> money on UPS. And if they are cheapskate enough to refuse, do you
> really
> want them as a customer?
It's not a problem of the framework as much as it is a problem of
frequent writing to the table. The move to views is a huge step in the
right direction. I had a client whose software was installed on their
client's systems, and thus they couldn't control things such as the
quality of electrical service, etc. They too were troubled by frequent
calls to deal with corrupt data as a result. My first change when I
came on board was to switch from direct table editing to views with
discreet updates back to the table. Remember, corruption can only occur
during a write operation. Having the data edited in memory cannot
corrupt the copy on disk. By switching to views, the window where
corruption could occur was reduced dramatically, and so did the
corruption suppport calls.
___/
/
__/
/
____/
Ed Leafe
http://leafe.com/
http://foxcentral.net
©2002 Ed Leafe |
