Re: Re: Disk partition names

I understand it's a sidai thing - there is no sidai email list yet.  And
there could be an alternative policy group that does it another way.

And I agree that user-friendly and user-hostile names are good.

I'm just wondering about the pros/cons of using "information hiding"
names for the user-hostile disk names.

For exmple, if one used ._diskN names and a disk gets
added/deleted/reorganized, it will take Extra Effort to see what things
should be renamed.

OTOH, unless there is an "automatic partition name function", there will
be a different level of effort on some OSes when this happens.

I was just wondering if we could discuss this a bit.

H
---
> Harlan asks:
> 
> > I'm reading blessed-names.txt and I'm curious.
> > 
> > Why ._disk1 ... ._diskN instead of things like:
> > 
> > 	._c1t0d0s3	(Solaris uses cNtNdNsN style partition names)
> > 	._vg01_lvol3	(HP-UX uses /dev/vgN/lvolN style partition names)
> > 	._hd9		(AIX uses /dev/hdN style names)
> > 	._da0s1e	(FreeBSD uses /dev/daNsNX style names)
> 
> [NB: these names are *not* an ARK (mechanism) thing -- you
> can name your partitions anything you like in Arusha land.
> These names are a sidai (one possible policy...) thing --
> it's how Matt and I do things...]
> 
> No deep reason at all.  The only thing is to have *some*
> kind of name that tells users that "here be dragons".
> 
> If, instead, users spot relatively friendly names,
> e.g. /export1, you will eventually find these names wired
> into PATHs and who knows what else, and you lose the Very
> Important ability to move things around among disks as
> needed.