Bad performance using ncp mounts under Linux, using ncpfs

Hi there, we are in process of migrating a few NFS mounts from one server
to NCP. We want to set the same protocol for accessing our file system
in all of our servers.

But, we are discovering that the performance using NCP is poor thant using NFS.
For example, accessing a 100 MB file for reading, takes about 3,5 minutes, instead
of 0.9 using NFS.

Using Windows for copy the same file takes about the same time than using nfs. So,
at this point, we have use Wireshark to see whats happening, and we see a lot of
read/ok messages, because the maximum data block size read is 1024. IN windows, this
is not the behaviour. If the reads are slow, the write opperations are... you know :)

So, I dont know if we can set differents block size while we are accessing ncp mount, because
AFAIK it seems that is the problem...

thank you so much!!