Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 1 Sep 2015 01:14:12 +0200
From:      Polytropon <freebsd@edvax.de>
To:        Quartz <quartz@sneakertech.com>
Cc:        freebsd-questions <freebsd-questions@freebsd.org>
Subject:   Re: Stop using a SATA drive
Message-ID:  <20150901011412.ec085f86.freebsd@edvax.de>
In-Reply-To: <55E45757.9000901@sneakertech.com>
References:  <CAPi0psvT5aaHR7kU%2B28qwVDdutyMn7LjhFUGZRWctz4gGfgvgw@mail.gmail.com> <20150824214252.53aa04c6.freebsd@edvax.de> <55DEF869.1010202@sneakertech.com> <55DEFB5A.3080408@FreeBSD.org> <55DEFC74.3040609@sneakertech.com> <20150828000602.b9a288a8.freebsd@edvax.de> <CAPi0psvMcrHKCQK9kBSacMmNie_042q9RQtKit6k4dvwA0GJQg@mail.gmail.com> <20150829220809.438bbf30.freebsd@edvax.de> <55E45757.9000901@sneakertech.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, 31 Aug 2015 09:32:07 -0400, Quartz wrote:
> > The remaining question is: Is it technically valid to
> > remove a device special file from the devfs file system
> > corresponding to a device that currently is not in use
> > (anymore), but is _present_ (attached to the system in
> > some way)?
> 
> I keep using OSX as a point of reference, but the way they do it is that 
> once the drive has been "ejected", it's effectively not present anymore. 
> Their mental model is that a drive can be physically attached without 
> being 'connected' software-wise, just that the process of establishing 
> that connection when a device is plugged in has been automated.

This leaves the more or less philosophical question: If the
user intends to use the disk he just ejected (but which is
still connected) again, what does he have to do? Obviously,
he cannot mount it anymore - no device file. Does he need
to pull the USB cable and so powering down the device, and
then plugging it in again, causing a somehow superfluous
power cycle, to make the device file re-appear, or is there
some "scan for new devices" (equivalent to "camcontrol reset")?



> Personally I've never had a problem with this mental model.

The problems start when you leave the predefined path. :-)



> Many 
> different things can be physically plugged into a computer without 
> actually functioning (ie; network cable) so I don't see why drives 
> should have special rules.

This is in fact correct, and again philosophically, think
about the importance of the _time_: Just because I plug
in something _now_ - it doesn't imply that I intend to
interact with it right away, or even worse, in the way
the OS believes I will.

Example 1: I have a hard disk that is to be subject to a
forensic analysis. What I do _not_ want is that it will
be automacially mounted r/w, maybe searched for files,
or written to.

Example 2: I have a DVD which I want to copy some files
from. Not now, later. What I do _not_ want is that it
starts auto-playing.

Example 3: I have a PCCARD wireless network adapter which
I need the next day for setting up a WLAN AP "man in the
middle" for traffic diagnostics. What I do _not_ want is
that it automatically connects to my neighbor's open WLAN
right now, and phone home.

The "problem" with automatism is that one size doesn't
fit everyone. :-)



> Any anyway, when I eject a drive it's because 
> I'm about to physically remove it.

In this, and _only_ in this kind of context, it's a fully
valid approach. It's just that the question of the technical
implementation and its valididy remains, and maybe even that
is just a matter of taste. :-)



-- 
Polytropon
Magdeburg, Germany
Happy FreeBSD user since 4.0
Andra moi ennepe, Mousa, ...



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20150901011412.ec085f86.freebsd>