John, your problems are very similar to ones we had. It took a lot of arguing, but the "network guru" was finally sent to California from their east coast headquarters - he found the network server was so overloaded that periodically it would skip a command. In our case, it was skipping append blanks and memo updates just like you're seeing. We sweat blood trying to bullet-proof the append blanks - finally ended up checking each new record to see if it was truly there. With heavy appends this added noticeably to the processing time (although less than the downtime of missing records and system crashes when a record couldn't be found).

They offloaded a heavy printload to a separate server and gave us 100baseT cabling and new hubs and all the errors went away.

HTH
Barbara

>hi Jim
>
>i have an app running at 28 different locations. all but one location uses intel cpu's and it runs 24/7/365 without any problems whatever. the other one uses machines with k2 450 cpu's. on that system (4 machines in an office) we have all sorts of things happening that i can't get replicated on any intel cpu. it is mainly data loss in memo fields. they will enter stuff into a memo field and next time they need it, it is just gone. every now and then when appending a new record it too will dissappear. i am very sure of the code (been programming for 15years and know how to append a blank by now), so something else has to be the problem. if you get to the bottom of it, please make sure you post it on UT!!
>
>good luck - john