[Libburn] Re: Yes, I'll be working on it over the summer

[scdbackup at gmx.net]

Hi,

Derek Foreman wrote:
> Yes, I'll be working on it over the summer.

(bouncing hopefully)

Will you look at the two big itches of cdrskin ?

- The lack of TAO writing with nonpredicted length.
  I got some (sparse) feedback and all complained that they
  cannot simply pipe a data stream to CD.
  The current workaround in cdrskin is cumbersome and time
  consuming (i.e. to announce full 650 MB and to have cdrskin
  padding up what's missing).

- The lack of an official method to avoid a full device scan
  but to touch only one single device via a persistent address
  (like /dev/sgN or /dev/hdX).
  Best motivation might be a system with two burners where the
  scan will stall while one of them is busy.
  Again i propose the whitelist method which is in use by
  cdrskin since half a year now. It is reliable, orthogonal
  to any deeper entrails of libburn, easily extensible to new
  device types, ... (well, enough of advertising already).

Then there are smaller itches like the fact that cdrskin has to
fork and start programm eject in order to get the tray out.

I'm standing by :))
My own test version of cdrskin is based on a libburn CVS version
from february 2006. Just the whitelist thingee is still non-standard.


Have a nice day :)

Thomas