>
> Ferhat, your  note is  very funny  indeed but does  not have  much useful
> suggestions in it. I'm the one who wrote the line monitor and I believe I
> know VM pretty well. The reason it monitors records and not bytes is that
 
  I don't know VM at all. Neither want to learn any more. Huh.
 
  My micro suggestion was 'monitoring according to byte count' which you
 say that it is not possible for RSCS.
 
    text cut....
> the file, because  a file containing records of, say,  132 times the same
> character  would take  very  little  time to  transfer  (because of  data
> compression).
 
  Data compression is another thing. Monitoring can be (sorry, could be)
  done before the compression.
 
>
> So,  it is  true that  monitoring  on the  number of  records is  stupid.
> Unfortunately  this is  the best  that we  can do,  and it  is definitely
> better than  nothing: in  practice, there are  *not many*  50,000 records
> files  that contain  only 3  characters  per line,  and *a  lot* of  them
> containing FORTRAN programs. The 5-10 users in the world who like to send
> 50,000 records of 3 characters will just have to wait, and I think that's
> better  than making  tens  of thousands  of people  wait  for their  mail
> because large files are being sent.
>
>   Eric
 
  It is sad not to be able to access to a really needed information. RSCS
dependence seems bad. Well... You are right. Thank you.
 
Ferhat