Subject: | |
From: | |
Reply To: | |
Date: | Sat, 4 Nov 1995 19:53:26 -0600 |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
According to Ross Scroggs:
>
[snip details of question]
>
> Interestingly, if you actually perform the build (-200,10), the record size
> is shown as 2018. Hmmm, where does that come from? Well, the file system says:
> if we publish the number 2022 as the record size, someone might actually try to
> write a record that big and it will fail! Why? Well, 2 bytes for length,
> 2022 bytes of data and 2 bytes for end-of-block terminator equals 2026, oops,
> 4 bytes two many. So the file system publishes the record size as four bytes
> less than the block size just in case you want to write that mega record.
> BUT!!! The file system only performs the subtract 4 business if the blocking
> factor is greater than 1. If you perform the build (-2022,1), the record size
> is shown as 2022 and the block size is 2026 so you really can write the mega
> record.
This is a great example of how MPE programming protects us from ourselves and
why we love hp3000s. Of course, one could argue that MPE just confuses us and
obscures what is going on, but there there is this mailing list to get
explanations from....
--
-- - - - Speaking for myself and not necessarily anybody else - - - - - -
Richard Gambrell | Internet: [log in to unmask]
Mgr. Tech. Services | POT: 504-483-7454 FAX: 504-482-1561
Xavier University of LA | Smail: 7325 Palmetto, New Orleans, LA 70125
|
|
|