Glenn,

I have been following some similar (I think) problems for a while now.

One thing that seems to appear in a few of them is that a table will open correctly bu t its CDX will not, though *NO* error is reported. I am even less sure, but I think this has occurred on NT systems (network part, that is).

If you can devise some way to double-check the status of the CDX immediately after Open, it could provide a lead which might then let you devise some kind of "bypass".

All this is, of course, highly speculative.

Good luck,
Jim N

>I have been working with Visual FoxPro 5.0 for several years as a reseller and as a programer. Our customers seem to have a very high incidence of files becoming corrupt. For example customers repeatedly get "Index Tag Not Found " errors. Our current problem is a real handful. We have a customer who backs up his data and then attempts to close a period. During the close the program crashes so he will restore his data and try again. The second time the program works fine. Even though the data was exactly the same. At first I thought that the customer was doing something silly to cause this situation, so we downloaded his complete data set and we were able to reproduce the error on our computers. The thing that complicates this situation is that the problems can appear and disappear at will.
>(Validate Database Ok Index All OK) Specifically the code:
>
>USE APCHCK
>COPY TO TEMP1 &&/// CRASHES "READ ERROR IN FILE"
>* OR
>USE APCHCK
>DO WHILE NOT EOF()
> ? CCHKNO
> */// Crashes with same error at records 981,1472,1965,2468,2951
> */// Approximately every 492 Records
> */// Repeatedly crashes then repeatedly works even though the files are identical
> SKIP
>ENDDO
>
>USE GLUNJE
>DO WHILE NOT EOF()
> ? CACCTNO
> */// Crashes at records 952,1430,1908,2387,2865,3344,3822,4300, */// 4779,5257,5735,6214
> */// Approximately every 478 Records
> */// Repeatedly crashes then repeatedly works even though the files are identical
> SKIP
>ENDDO
>I ran into a similar situation with several years ago with a bug in Clipper 87 indexes. But that error can be consistently reproduced on any computer 100% of the time. This error is intermittent. (crash,crash,work,crash,crash,crash,crash,work,work,work,work,crash,work)
>
>Any suggestions would be greatly appreciated. Thanks in advance.