[kwlug-disc] So why not tar -cf tarball.tar a.xz b.xz c.xz, instead of tar -cJf tarball.tar.xz a b c ?

[B.S.]

On 10/23/2016 04:30 AM, Chris Irwin wrote:
> On Fri, Oct 21, 2016 at 06:09:44PM -0400, B.S. wrote:
.
.
.
> It isn't prevalent because it isn't practical.

Of course it is. (And any backup centric person would deem it a 
requirement - how do you test the backup, or determine how much of a 
degraded backup is trustworthy?)

> But note that those are just shorcuts for convenience. ...

I appreciate all that you have said.
(Snipped for brevity, but lots of very good stuff.)


tar -cf mytar.tar --in-command="gzip @" -
(if --in-command existed)

or (mixing OS') gzip @filelist | tar -cf mytar.tar -

- Sadly, losing filesytem info in the process, like filenames and metadata.

Unlike 'gzip @filelist ; tar -cf mytar.tar @filelist'

Would be just as practical as the rest.

>> The question is: With a damaged tar file, and the day -will- come, how
>> do you know which files within what remains are good and which ones
>> are damaged? (It's unlikely to be all of them.)
>
> You don't. You throw the damaged file away, then you restore a good copy
> from a backup.

<sigh>

The scenario I'm envisioning is that this IS the backup file from long 
term storage, that has degraded over time.

- which, for anyone who never used magtape (tar = Tape ARchive, after 
all), sometimes seemed like 'long term' was a whole 2 weeks ago.


I do understand / appreciate the ecosystem - just curious that 
individual file validation, especially if tar's have any level of 
fragility at all, isn't much mentioned.