[Libburn] Thesis on Libburn

[Derek Foreman]

On Wed, 31 Mar 2004, Steven Van Impe wrote:

> > 1) The project is still in an early stage, so the code is still
> > comprehensible ;)
>
> that makes it easier indeed
>
> > 2) The project is written in pure C and uses standard C libraries.  Most
> > current free software projects are either based on GTK+ or QT, and while
> > these are both excellent toolkits, if you haven't done much C
> > programming to begin with, learning a toolkit and a language at the same
> > time can be daunting.
>
> that's one of the reasons I chose this project
>
> > My recommendation would be to build the latest CVS code and start
> > banging on it.  Find something that doesn't work the way to expect it
> > should, and fix it :)
>
> that's not the approach I had in mind :)
> what I will need to know is exactly what I need to learn on cd recording, so I
> can do some serious studying on that
> my teachers will probably be able to assist me in finding good material on
> that, if it isn't available on the internet
> also, I'll spend the first part of my summer 'holiday' doing some smaller
> exercises in C, finishing 'the C programming language' and taking a look at
> the POSIX API
> when all that preparation is done (should be by the end of the summer I hope),
> I can get to working on the code, mostly small stuff in the beginning, while
> I get used to working on a free software project, but I do hope to make some
> bigger contributions during the remainder of the year (which I think is
> possible, since I'll have lots of time, almost every day, to work on this
> project)
>
> I guess now I'll wait and see what the project maintainers think of all this,
> and if they're ok with it, I'll propose this as a thesis subject to my
> teachers

Hi.  I'll be your maintainer for today.

For documentation, there's a great link at the bottom of the libburn home
page.  (the link is http://le-hacker.org/dvd.html)

Especially important are inf 8090, ecma-130, and ecma-119 (iso 9660).
Also, a copy of the mmc docs would be good.

Also, some of the forums on cdfreaks.com are quite technical.

The "Red Book" standard is not freely available, but it might be in your
university's library.

Also, there's a linux sg programming guide that's quite good if you're
interested in those parts, it still applies to the SG_IO stuff for 2.6
kernels.

And I can provide at least a little help (to anyone interested in working
on libburn).



Right now libburn needs a lot of work.

I'm not sure of the status of libisofs, Todd and Ben know that part of the
tree better than I.  I think it's capable of writing an iso9660 filesytem
with no extensions (8.3 filenames only).  so joliet and rock ridge would
be nice.  and el torito, and all that jazz.

Libburn itself needs to interface nicely with HAL (but not rely on it, so
we can still support architectures that HAL doesn't).  I'd also like to
see it ported to a few other systems, especially win32. :)

It doesn't do TAO recording yet, but it's close.  So by the time you're
ready to get involved, that won't be an issue.

It doesn't properly handle indices.

It doesn't read discs at all yet - that alone is an absolutely immense
area of work.

It doesn't handle multi-session discs.  The API is supposed to handle
this, in theory, but the code behind it hasn't been written.



Are you allowed to hand over the copyright on your code, or does your
university expect to keep it?

There's certainly a lot left to do, and if you can get course credit for
helping out, that's cool with me. :)