[Info-vax] HP wins Oracle Itanium case

[Bob Koehler]

In article <3O4Zr.16459$NK4.8611 at fx24.am4>, ChrisQ <meru at devnull.com> writes:
> 
> In the end though, what get's stored on disk is a stream of bytes at the 
> lowest
> level.

   That's one point of view.  Another is a collection of blocks, or a
   stream of bits.  Any of which may or may not meet the programmer's need.

> To make the most flexible use of that, an interface to that 
> stream should
> be provided, even if there are other structures eg: rms, above. Someone else
> said that the qio interface could be used, but seem to remember that as 
> being
> fairly arcane at the time.

   Why should a programmer have to deal with the meta data that RMS
   hides?  RMS does provide three stream files already, or "undefined"
   if you need even more flexibility.

   When I program in HLL, I don't want to deal with CR vs. CRLF vs. LF
   vs. leading count vs. indexed vs. ...   I just want to get the data
   in a form understood by my HLL's programing paradigm.  I spend too 
   much time now dealing with CR vs. CRLF in a mixed UNIX/Windows
   environment.  On VMS, I just ignore that most of the time.