FU Logo
  • Startseite
  • Kontakt
  • Impressum
  • Home
  • Listenauswahl
  • Anleitungen

Re: [linux-minidisc] introduce recorder/filesystem concept split; move autodetection to libhimd

<-- thread
<-- date
  • From: Michael Karcher <Michael.Karcher@fu-berlin.de>
  • To: manner.moe@gmx.de
  • Date: Tue, 30 Nov 2010 19:03:35 +0100
  • Cc: linux-minidisc@lists.fu-berlin.de
  • Subject: Re: [linux-minidisc] introduce recorder/filesystem concept split; move autodetection to libhimd

Am Sonntag, den 28.11.2010, 21:35 +0100 schrieb manner.moe@gmx.de:
> >My suggestion is to have a two-tier system - recorders and media. First
> >we detect recorders, that have (at least) the following properties:
> > - physical path (/dev/sdb on Linux, either E:\ or \\.\PHYSICALDRIVE3 on
> >Windows, whatever we need for libscg.)
> 
> There is no windows driveletter support in libscg.

In my oppinion, we should introduce drive letter support into libscg for
windows. It would be an easy patch, but it is against Joerg Schilling's
philosophy of "do it like Solaris does" which boils down to the
bus:target:lun triplet.

On Windows, libscg uses one of two different interfaces: SCSI
passthrough or ASPI (another cross-platform SCSI API, provided by
aspi32.sys). I guess that for the Hi-MD stuff we want to use the native
and more modern SCSI Passthrough Interface (SPTI). This interface is
directly bound to drive letters! But libscg takes a lot of effort to
hide it and squeeze it into the bus/target/lun abstraction.

> On Windows we need the physical path in the form <scsibus>:<target>:<lun>.

"physical" is a quite euphemistic word for it :), it's more like
virtual. Windows internally has some adapter/bus/target/lun layer for
adapters providing real SCSI busses with multiple possible targets, but
that layer is not used at all for USB or FireWire mass storage devices
that are accessed using the SCSI command set. libscg collects all busses
of the different SCSI host adapters and all drives with SCSI passthrough
capabilities into one big list and then assigns bus numbers to all those
devices.

> We should iterate over all scsibusses, targets and luns to find himd devices 
> by vendor and product ids as connected himd devices have no driveletter 
> assignment if no medium is present at connection.

That last claim is not really true. If I connect the MZ-RH1 in Hi-MD
mode (you can set the default mode in the menu) without any medium to a
Windows XP computer, it gets a drive letter immediately. In fact, libscg
won't even work until there is a drive letter assigned, except if you
use it in ASPI mode (which you most likely didn't). Of course, if the
default mode of the MZ-RH1 is set to "MD" instead of "Hi-MD", you
*don't* get a drive letter, but this is because in that case a NetMD
AV/C-over-USB device will be connected instead of a USB mass storage
device. If a Hi-MD medium is inserted, the device will always connect as
mass storage device, even if the default mode is set to "MD".

Regards,
  Michael Karcher




<-- thread
<-- date
  • References:
    • Re: [linux-minidisc] introduce recorder/filesystem concept split; move autodetection to libhimd
      • From: manner.moe@gmx.de
  • linux-minidisc - November 2010 - Archives indexes sorted by:
    [ thread ] [ subject ] [ author ] [ date ]
  • Complete archive of the linux-minidisc mailing list
  • More info on this list...

Hilfe

  • FAQ
  • Dienstbeschreibung
  • ZEDAT Beratung
  • postmaster@lists.fu-berlin.de

Service-Navigation

  • Startseite
  • Listenauswahl

Einrichtung Mailingliste

  • ZEDAT-Portal
  • Mailinglisten Portal