| 1 |
On Wednesday 23 April 2008, Mark Knecht wrote: |
| 2 |
> On Wed, Apr 23, 2008 at 12:40 AM, Alan McKinnon |
| 3 |
|
| 4 |
> > Personally, I prefer labels over other disk id methods. I get to |
| 5 |
> > choose the label myself and can ensure they are unique in my world |
| 6 |
> > (but maybe not in the universe like UUIDs are). If I have to mkfs a |
| 7 |
> > volume from scratch for some reason, it's easier for me to to |
| 8 |
> > re-use the same label than to re-use or copy-paste those long UUID |
| 9 |
> > strings |
| 10 |
|
| 11 |
> I like labels also. I've had a couple of cases where I've taken a |
| 12 |
> drive out of an old system but kept the drive around. Later I put the |
| 13 |
> drive in a 1394 drive case.I checked the drive label and immediately |
| 14 |
> knew it was a drive with ripped music, sessions I've recorded in |
| 15 |
> Ardour, etc. Labels are human readable and I tend to make them quite |
| 16 |
> descriptive. |
| 17 |
|
| 18 |
Just to expand a bit: UUIDs are guaranteed to be unique in the whole |
| 19 |
world, that's why distro installers use them - you can issue guarantees |
| 20 |
that the installer will get it right. |
| 21 |
|
| 22 |
LABELs have no such guarantee and installers need the user to be clued |
| 23 |
up enough to pick decent ones. As we all know, average users often |
| 24 |
aren't up to that. The majority of Ubuntu's target market (just to pick |
| 25 |
a random example) certainly aren't. |
| 26 |
|
| 27 |
So the installer is between a rock and a hard place - use the method |
| 28 |
guaranteed to work today, but is not really human-readable, or use a |
| 29 |
lesser method with a few caveats (like a trained user). It's the old |
| 30 |
story all over again - use the one that works best for you as long as |
| 31 |
you know enough to be able to decide |
| 32 |
|
| 33 |
-- |
| 34 |
Alan McKinnon |
| 35 |
alan dot mckinnon at gmail dot com |
| 36 |
|
| 37 |
-- |
| 38 |
gentoo-user@l.g.o mailing list |
Archives