Provide an easier to travers data type in xml_state for the
user and groups. The get_users() method returns a list of
tuples which assigns each user list the group name and id
it belongs to
try normal mount first, if this does not work try fuse mount
Normally this is done automatically by the mount program but
if not we will give the direct fuse mount a chance
In addition to fat also exfat is now supported as persistent
write filesystem. Because of the limitations of a fat filesystem
fat and exfat are only used as a container filesystem providing
an ext4 linux filesystem as a cowfile. The algorithm to create
the size of the cow file has also been changed to use half of
the size of the write partition or on fat a max size of 4G.
The size of the cowfile is also prepared to become overwritten.
However the XML definition and implementation to do this
is still missing
The hybrid write partition is created via fdisk. Some version
of fdisk does not send the ioctl to let the kernel reread the
table or fdisk itself holds it busy. Thus we actively initiate
a reread via blockdev
For overlay filesystems not writing into a tmpfs performance is
more important than safety. We use this combination of options
for now, if you encounter stability problems please let us know
Instead of a fixed name 'cowfile' we allow a custom name which is
predefined in HYBRID_PERSISTENT_FILENAME and prepared to become
overwritten by an XML defintion whose implementation will follow
later. Reason for the change is that a cowfile is visible as
plain data file to the operating system if e.g used on a live
stick. It should be more clear to the user what this file is
good for
The methods did not cope well with filenames containing bash
characters with special meaning e.g spaces. For use with a
iso hybrid cowfile whose name is visible in the OS the methods
should be able to work with any given filename
A change in kiwi/schema/kiwi.rnc triggers the build of the
schema rng version as well as an auto update of the data
structures via python's generateDS. The intermediate xsd
schema format is only relevant for the data structure
creation process
At the moment there is no class to build the pxe client config
file from information provided with the optional pxedeploy
section. However this is not fatal to the build because the file
could be created manually too. Thus a warning message is shown
which will go away when the pxe client config file creation
has been ported
Put the creation of a result object into the builder classes
and return them from there. The builder instances knows about
the results not the task instances
If disk format and install media is configured together only
the install media will be built and a warning message for
skipping the disk format is shown
