Date:         Fri, 12 Jul 2002 11:21:12 -0400
Reply-To:     Tom Frenkel <taf2@NYU.EDU>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Tom Frenkel <taf2@NYU.EDU>
Subject:      Re: PROC COPY and checksums
In-Reply-To:  <200207111917.g6BJHeo06971@listserv.cc.uga.edu>
Content-Type: TEXT/PLAIN; charset=US-ASCII

Jordan --

I did a PROC COPY on my Win2K PC (SAS 8.0) and, using my trusty "V"
browser (shareware from http://fileviewer.com/ ) in Hex mode, could see
that some (unprintable) characters in the headers were different in the
two SAS datasets.

Perhaps you might want to try a literal (binary) copy and see if the
results still differ?  That would of course indicate a network problem.
HTH,

        --Tom
---------------------------------------------------
Tom Frenkel                     email: taf2@nyu.edu
Center for Health & Public Service Research
New York University
http://www.nyu.edu/wagner/chpsr/
---------------------------------------------------

On Thu, 11 Jul 2002, Jordan Hiller wrote:

> I'm having a problem copying large SAS datasets (~1G) across a Windows
> network.  The copied files seem to be corrupted, and I want to check
> whether some sort of network problem is screwing up the transfer.
>
> I have a DOS utility that will calculate an md5 checksum on the original
> file and the copy.  When I run this utility, the checksums aren't matching.
>
> So my question is this.  Does the checksum failure indicate that the files
> were corrupted on copying, or could the files be different because I used
> proc copy to move them? I don't understand too well what happens under the
> hood with SAS data sets -- is there some header information that could be
> changed as a result of using proc copy?