[Info-vax] HP wins Oracle Itanium case

[Paul Sture]

On Tue, 28 Aug 2012 11:58:35 -0500, Bob Koehler wrote:

> In article <k1h7li$r54$1 at Iltempo.Update.UU.SE>, Johnny Billquist
> <bqt at softjar.se> writes:
>> 
>> Ah. Misunderstood you. I thought I were using RMS for something more
>> advanced than just stream I/O. Yes, pretty much any language, under any
>> OS, will offer that much functionality. Using RMS, or whatever I/O
>> functionality offered by the OS.
> 
>    Yes, what you get with the way HLL on VMS use RMS is the ability to
>    exchange data between HLL, even though one is using a byte stream
>    paradimn and the other records, without having to deal with the meta
>    data.

It gets even better with a set of RMS wrapper routines, for you are no 
longer limited to the restrictions of a given HLL.  For example back in 
the early 80s the COBOL standard in force then said that you could not 
have two index keys starting on the same boundary.  While I could 
understand where that was coming from it was useless in reading files 
written by other HLLs.

>    Blows away folks when I access a keyed-indexed file "sequentially".

Again back to COBOL.  It would see an indexed file and say something like 
"Define the indexes to match what the file declares itself as".

Nuts to that. TYPE, EDIT, SEARCH et al can read an RMS indexed file, and 
while you might need to use CONVERT to get it back into its indexed 
format you can do lots there.

And yes, on a couple of occasions  I did use those utilities to get into 
the SYSTEM account when the normal guy was off sick and something needed 
doing.

And restored the original SYSUAF so he could carry on as normal. :-)

-- 
Paul Sture