🚀 go-pugleaf

In article <michaelh.717718108@Xenon.Stanford.EDU>, michaelh@Xenon.Stanford.EDU (Mike Hennahane) writes...
>sandrov@vxcrna.cern.ch (VENTURA,SAND./EP) writes:
> 
>>I'm experiencing some troubles with the Receive Record command:
>>I can't avoid it to create a new record each time it executes, despite what
>>stated on the manual which says that a record can be received in the current
>>record.
> 
>the manual says that you must create a *new* record for the file
>before you do a 'receive record.' i don't think that you can receive a
>record into the current record.

In fact the example in the manual shows how to receive by creating a new record.
But the description of the command (I'm one of the lucky men which works
with a reference manual in french, while scripting in english :-)
states "Si vous recevez cette fiche dans une fiche existant deja, les
anciennes valeurs des rubriques seront remplacees par les valeurs recues 
(i.e. If you receive a record into an existing record, old field values
will be replaced by received ones)". Which seems to be doubtless! Moreover, in 
the multiuser section there is a warning about being sure to work on a 
non-locked current record when receiving...
> 
>>What am I missing? Any help or more general suggestions on how to transfer
>>records between separate databases?
> 
>are you transferring new records, updating existing records, or both?
> 
>--mike
I'm keeping in sync a database running on non-networked machines. So I have to
receive new records, update old ones or delete some of them, which I try to do 
by the means of a file which contains the variables and records needed for this
task. Transferring new records and deleting works fine, updating existing ones
could be done by searching the record, deleting it and receiving the updated
version. But this scramble a little bit the "sequence number" function, on
which other scripts of the database rely. Ok, this can be fixed anyway, but the
sake of knowledge (or the lazyness...) made me wishing to understand why the 
receive command didn't work as described (and as wished)...

Sandro