problem with generic ascii product

[Dave Brown <dbrown@xxxxxxxxxxxx>]

Hi,

Some users here have encountered a problem with the generic ascii product
when attempting to fetch large subsets of some daily historical data.
The header at the beginning of the data does not reflect the full extent
of their request, or, it seems, the full extent of the data that is
actually downloaded into the browser. From examing the GenericLAS_debug.txt
it appears that if more than a certain maximum amount of data is 
requested (500000 words for this server), ferret enters a loop that
processes the data in chunks. It appears to me that the data header as
sent to the browser only reflects the data in the first chunk. 

As far as I can tell, it seems that ferret is generating the header. 
Is this a known problem, and if so, is there a workaround?
 -dave brown