Hi Bela,

I honestly do not know what is making it work. All I can tell you is that we had thought we were stuck when this was tried for lack of anything else left to try and, to our surprise, it worked.

Cheers,

Jim N

>Hi Jim!
>
>I mean, If I put a record lock manually, VFP will read the record from network, instead of its internal cache.
>I tested your idea. Yes, It works. I need the absolutely latest contest of a table, as You. Till now, I did a Flock()+unlock. I heard, that it forces VFP to reread the contest of an opened table (into its internal cache), so subsequent read acces to any record of this table will give the latest data. Without manual file locking, record data can be a non up-to-date cached data.
>From your trick, I guess VFP reread the contest of a record or whole table (?), when record pointer moves. Is it or not?
>
>BB
>
>>Bela,
>>
>>I had good luck in a very particular circumstance... (in FPD 2.6) I had a 1-record table which I always needed the absolutely latest content of and found that a "GO 1" did the trick for me.
>>
>>However, I don't really understand what you want/need, since you say 'read" and 'without locking'. I never knew of a read access which did a lock. Also, you can legally lock the same record many times in an app (if that's really what you want to do <bg>).
>>
>>Can you give more details?
>>
>>Jim N
>>
>> >Hi All!
>>>
>>>Can anybody tell me, how to force VFP to reread an opened table from the network without locking it? I thought that SET REFRESH affects it, but it didnot...
>>>
>>>BB