Xref: utzoo comp.os.vms:9130 comp.unix.wizards:11476
Path: utzoo!utgpu!water!watmath!clyde!att!rutgers!gatech!uflorida!haven!ncifcrf!nlm-mcs!adm!smoke!gwyn
From: gwyn@smoke.ARPA (Doug Gwyn )
Newsgroups: comp.os.vms,comp.unix.wizards
Subject: Re: Records; VMS vs. UNIX file system
Message-ID: <8618@smoke.ARPA>
Date: 3 Oct 88 08:03:44 GMT
References: <411@marob.MASA.COM> <3597@encore.UUCP> <3438@crash.cts.com> <7296@rpp386.Dallas.TX.US> <13800@mimsy.UUCP>
Reply-To: gwyn@brl.arpa (Doug Gwyn (VLD/VMB) <gwyn>)
Organization: Ballistic Research Lab (BRL), APG, MD.
Lines: 14

In article <13800@mimsy.UUCP> chris@mimsy.UUCP (Chris Torek) writes:
>In summary: the major difference is that Unix `records' are in the eye
>of the beholder, and not (as in VMS) supplied as any part of the
>system.

Chris's summary was pretty good, but in case anyone wasn't aware, he
was describing *disk files*.  Some of the more general notions of
"file" in UNIX really do have records, magtape and terminal input
being the obvious ones.  Of course the record structure is forced by
the nature of these devices, so it isn't a design botch, but it is
something one should be aware of.  I recall not long ago finding out
much to my annoyance that at least one version of "cat" was losing
record size information due to having been converted to stdio instead
of read/write.  And yes, it DID matter.