On 05/24/2016 03:40 AM, Alan Altmark wrote:
On Monday, 05/23/2016 at 08:30 GMT, WF Konynenberg <[email protected]>
wrote:
My concern was that Alan Altmark suggested that a critical bit of
information to positively identify these files was not available to
Linux programs, which is apparently not the case.
I meant that the attributes (user.record_format and user.record_lrecl)
aren't expressed by the driver as a "normal" file attribute that a
"normal" Linux application might be expected to extract on any file. I
assumed (but did not verify!) that if you copied the file from the CMS
disk to a native Linux file, the extended attributes would be lost. (I
don't know if extended attributes are a run-time conjuring of the file
system driver or if, once created, they persist with the file forever,
even if copied to a non-cmsfs-fuse file system.)
If they do persist, then a cmspack utility would work as well as CMS
COPYFILE.
Whether such extended attributes are copied along with the file depends
on a) whether the target file system supports extended attributes (most
Linux file systems do), and b) whether the user asked for them to be
copied. By default, the common cp command will not copy extended
attributes. You have to ask for that to happen. The simplest way is
perhaps to do "cp -a ..." to invoke archive mode. Other copying tools
may or may not be a bit more careful to by default copy relevant
metadata along.
Willy
----------------------------------------------------------------------
For LINUX-390 subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
----------------------------------------------------------------------
For more information on Linux on System z, visit
http://wiki.linuxvm.org/
