[game_preservation] Dreamcast Archival

[Mike Melanson]

>  From my research, and specifically for the Dreamcast, there is such a

> method already available using a Broadband Adapter and a Dreamcast

> hombrew application called httpd-ack.  Apparently there are plenty of

> tutorials online.  Might I also remind everybody that data migration,

> even for preservation, altruistic or benevolent purposes, certainly

> borders on the darker end of the gray area.  Non-profit and educational

> institutions DO get some leeway for fair-use and archival purposes but

> it's not enough to legally justify copying such material as far as I can

> tell.  Don't take me for a legalese expert.  I recommend consulting with

> a lawyer before taking any actions.


Legality noted; I'm mostly of the "privately archive it now rather than
regretting it later after the media has deteriorated" school of thought
(would this be the "Jason Scott" method? :) ).

I could go the broadband adapter route, rendering all of this academic.
Then again, I'm partially in this for the academic challenge. I could also
go the uncompressed serial line route and just devote up to 28 hours to
archiving each disc (I've done it before). The nice thing is that it
doesn't tie up any other computing resources.


> Going further.  Even with a .cue sheet and RAW data/audio .BIN dumps of

> each track, there is still sub-channel data missing and quite a few

> other extraneous and un-captured bits as well.  True preservation, in my

> mind, encompasses the entire disc, bit for bit.  I've had an idea on the

> back burner(crude cd-r joke?) for some time now conceptualizing how to

> go about creating a file system/disc format agnostic method to copy data

> off of any optical medium given the right hardware.  From my

> understanding, all that one has to do is gain low level hardware access

> to an optical drive, start at offset -1000 (give or take) and start

> reading the disc's peaks and valleys.  Correct me if I'm wrong, but

> wouldn't it be relatively simple to code such a tool for a specific

> drive if you were familiar with the specific drive's I/O protocols?

> Alternatively, you could always just scan a disc's data side in at

> 4000dpi and have a virtual laser read the peaks and valleys, right?

> KIDDING!  Kinda.  Double layer discs would be a bit difficult to scan

> anyway.


Yeah, there is a lot of thought and trade-offs that can go into the
archival process. And if you obsess too much, you'll never get anything
accomplished (if, indeed, your goal is to get something done).

When you mention "low level hardware access to an optical drive," you're
referring to the firmware level instead of the host OS level, right? Yeah,
that's a level deeper than I'l willing to go.


> So we've migrated, now what?  Optical drive emulation for data

> retrieval/referencing/usage of course!  It wouldn't be difficult to code

> a simple plug-in based tool for the new format agnostic disc image.

> Each file system type would need it's own plug-in to be readable though.


Here we come to trade-offs again. Whenever I discuss disc archival, I'm
always asked how I could take the archived data and reconstruct the
original disc. I'm usually at a loss at that point.

I guess it's a matter of motivation-- I've always been more interested in
ripping games open and studying how they operate, rather than actually
*playing* them. But that's just me.

--
    -Mike Melanson