FreeLists / openbeosstorage / [openbeosstorage] Re: ISO9660/cdrom progress

["Ingo Weinhold" <bonefish@xxxxxxxxxxxxxxx>]

> Tyler Dauwalder <tyler@xxxxxxxxxxxxx> wrote:
> > Thanks to some help from Marcus and Axel, I think I'm finally 
> > getting 
> > close to wringing some session info out of my cdrom drive. :-) I 
> > also 
> > have a nearly working ISO9660 fs module, and finishing it should be 
> > fairly straightforward. I'll post more news when I have some.
> 
> Very nice!

Yep. :-)

> We should also have priorities for the file system modules (i.e. a 
> UDF 
> file system should always get the disk, not the ISO-9660 file system, 
> if possible).

That shouldn't be hard. We can add a function returning the priority to 
the FS module interface.

> Also, how does the system deal with multi-session CDs (and the like)=
> 3F 

As it does right now, via the session modules.

> We might also want to have write support built-in,

If you mean, initializing a FS and writing partition tables, then 
that's what the original DriveSetup add-ons were doing. The only 
problem is, that both tasks need add-on-specific parameters, that in 
the case of the DS add-ons is retrieved from the user by popping up a 
dialog. The only way, I see, is to create a separate DS add-on API for 
this.

> disks that span over 
> several volumes, etc.

RAID? Then this should be done on a lower level, I think. Like a driver 
that publishes a respective device.

> At least we should think about every possible extension we might want 
> to add to that piece of code.

Well, yes, but there aren't that many coming to my mind. :-P

> > It appears the the module emulator is not working quite right in 
> > instances where you have two (and probably more) modules of the 
> > same 
> > type. When testing the iso9660 fs module (alongside the bfs 
> > module), 
> > both modules are init'd and uninit'd initially when the list is 
> > created, but only the first one (bfs, in this case) is returned by 
> > read=5Fnext=5Fmodule=5Fname(), which then returns an error the 
> > second time 
> > it's called. I'll look into it more if I get sessions working.

I should find some hours time today for analyzing and, hopefully, 
fixing the problem. I was wondering anyway, why this code, that I wrote 
from the the scratch, was working that quickly without much debugging. 
;-)

CU, Ingo