[DVDnav-discuss] dvdread patch for broken dvd's

[Nico Sabbi]

Il giorno mer, 08/10/2008 alle 21.10 +0200, jezz at hkfree.org ha scritto:
> On Út  7.říj, Nico Sabbi wrote:
> > Il giorno lun, 06/10/2008 alle 22.09 +0200, jezz at hkfree.org ha scritto:
> > > Hi,
> > > here are my answers. Cosmetics are not problem, but before fixing them, the idea
> > > of this code must be approved. For file there are 2 possible tag values - 261
> > > for File Entry (4/14.9 in Ecma 167) or Extended File Entry (4/14.17). For some
> > > VOBS on DVD there is no File Entry or Extended File Entry. But there is
> > > Extended Attribute Header Entry with tag 262 (4/14.10). If I handle 262 entry
> > > as 261, than DVD plays fine, but that is wrong - EA Header is not File Entry +
> > > some data. It contains data associated with some File Entry, but there is none.
> > > Therefor this DVD is broken. But ISO 9660 filesystem is ok - kernel mounts this
> > > DVD and totem can play it (totem don't plays discs directly, but only mounted).
> > > I can play VOB directly using mplayer, but I can't use dvdnav or some mplayer front-ends.
> > > This ISO stuff uses correct ISO fs if file search on UDF fails. This fixes my
> > > problem - and maybe others, too.
> > > 
> > 
> > I agree completely with your method: accepting tag 262 as valid is plain
> > wrong, and for the rest (without reading the code of iso9660) the code
> > is fine
> > Please, don't top post in mplayerhq mailing lists
> 
> The ISO part is programmed using this documentation:
> http://alumnus.caltech.edu/~pje/iso9660.html
> I suppose, that simplified ISO spec is sufficient for DVD video. Moreover this
> is only fallback method.
> 
> Is there any coding style for dvdnav/dvdread sources? May be GNU indent will
> fix this :). Or did you mean something different as cosmetics?
> 

cosmetics are non-semantic changes; they are fine as long as they are
isolated from semantic changes and as long as they don't change the
style of the enclosing files