Subject: | Re: tc.add stops working
| Date: | Fri, 29 Dec 2017 10:51:32 +1000
| From: | "Phil Hassid" <philhassid@ozemailDOTcomDOTau>
| Newsgroups: | pnews.paradox-development
|
Hi Mike,
Thanks for all of that. I will look into all your suggestions as well and
report back in the coming days.
Phil
"Michael Kennedy" <Info@KennedySoftware.ie> wrote in message
news:5a44ec71$1@pnews.thedbcommunity.com...
> Phil,
>
>> If it were possible, how would I even test for that and how would
>> I relocate it elsewhere?
>
> It's extremely unlikely that executable code is damaged (or unreadable),
> and the app still runs without throwing I-O errors. You could run a
> surface scan (CHKDSK, etc), but it might take a long time.
>
> Along the same lines, it's possible that the disk structures are damaged,
> and then the reading of a chunk of the executable/library is actually
> retrieving a chunk of something else, but that should also throw major
> errors! CHKDSK, etc, should flag all structural errors (quickly).
>
> If you have a backup of the executable files, just run a file-compare (FC,
> DIFF, etc), between a backup and the current ACTUAL executables. That'll
> also show, or eliminate, possible corruption. If you have no backup of the
> live executables, you could re-build them, but:
> - Sometimes, time stamps, etc, are embedded in executables, so these
> bytes might then show as "differences".
> - You might now be using a new version of a compiler, or some other
> updated libraries, and these would also result in different executables -
> even with no changes to the app.
>
> Separately, could you try a Query (to Insert all the records), instead of
> an ADD? Probably a little slower than ADD, but.....
>
> - Mike
|