>Interesting. It could be that there are or at one time were plans to natively support an RS, but it's not implemented. I remember hearing that all the Intellisense code is actually in VFP6, but isn't activated, so I guess it's possible that there are some RS things inside VFP. It's also possible that these messages were added in SP3 for the VFPCOM.DLL, but haven't been documented.
>
>FYI, these errors remain undocumented in VFP7 help file. Maybe the final release will have them.

Just out of curiousity, Craig (and because I don't have Beta 1 handy), does 2041 still have the typo?

>There are also some errors that don't generate error numbers.

I'd say "I hate those" but my programs don't have errors...undocumented features, but not errors.:-)


>>I'm in the process of designing a new general purpose error handler, and decided to programmatically generate a table contain the error number and message returned. The current documentation states that the last valid error number is 2035. However, at least in VFP 6.0 SP5, the following list of error numbers are apparently supported and appear to be connected in some cases with ADO record sets.

2036     The current Help file is empty.
>>
>>2038     Microsoft Visual FoxPro doesn't support and cannot
>>         open this format of recordset. It contains a recordset with a shape.
>>2039     Failed to create recordset.
>>2040     Field mismatch.
>>2041     <b>Recodset</b> does not exist.

No, folks that isn't a typo on my part. It's what MESSAGE() returns.< s >
>>
>>I haven't check any of the error numbers prior to 2036, so I can't comment if any new, undocumented ones exist there.