On Sun, 2007-02-25 at 12:30 +0000, Ben Hutchings wrote: > I looked through the code used for resizing and I believe I can see the > source of this bug.
Well, maybe not quite. ;-)
> do_resize() gets two constraints: the bounds of the neighbouring
> partitions, generated by snap_to_boundaries(), and one from the
> filesystem code, generated by ped_file_system_get_resize_constraint().
> Then it combines them using ped_constraint_intersect() and passes the
> combined constraint into ped_disk_set_partition_geom().
In fact the normal parted front-end will not resize any partition if it
can't "open" the filesystem. I'm not sure quite what this involves but
this operation is not supported for NTFS.
I realise now that d-i is using an entirely different front-end which I
will need to look at.
Ben.
--
Ben Hutchings
Power corrupts. Absolute power is kind of neat.
- John Lehman, Secretary of the US Navy 1981-1987
signature.asc
Description: This is a digitally signed message part
