As far as I know we do use ntfs-3g already - on what are you basing that we don’t ? ntfs-3g is required for write support of ntfs, and we do have full write support.
As for udisks-glue.conf - as the poor schlub who had to write 90% of that monstrosity I feel your pain and sense of confusion. udisks-glue’s configuration file is very poorly documented (one man page basically) and has a very limited syntax that doesn’t allow for multiple matches or fall through of matches.
This is why there is a massive amount of duplication in the file - we need the script to add samba shares for all variations of file systems etc, but a match entry, staggeringly, cannot match multiple filters! Which kind of makes it pointless having separate filter and match clauses in the conf file’s syntax IMHO, but it is what it is and we have to work with it…
There’s two issues that you’re probably running into - one is that if you add even one “invalid” option in the automount_options, then the partition just won’t mount at all. That’s why we have separate matches for optical-udf and optical-other for example - because for a UDF formatted disk we need to specify umask=0000 for optimal file permissions for the mount point, yet umask is an invalid mount option for an iso9660 disk causing the mount to fail…sigh.
The other thing you may be running into is that udisks-glue itself rejects a number of valid mount options for some file system types based on a list that is hard coded in the source code.
For an example of that issue, see this kludgey but expedient workaround for EXFAT:
https://github.com/osmc/osmc/blob/master/package/diskmount-osmc/files/sbin/mount.exfat-fuse
From memory, fmask and dmask are valid options for an exfat file system, that we need to use, but are explicitly blacklisted for exfat file systems in the udisks-glue source code so any attempt to use them as automount_options results in the partition failing to mount at all.
Yeah, we could fork udisks-glue, patch the source code to remove these restrictions and supply our only custom version of udisks-glue, but in the end we decided it wasn’t worth the effort when a small script and a dpkg diversion could solve the problem for now. (Every extra package you fork means more work to maintain your own version - we already fork connman, lircd, and eventlircd among others, for example)
So yeah, customising the automount options is a right PITA at the moment. Eventually we will replace udisks-glue for something else more flexible, but that won’t be until well into the future, partly because Kodi has good support for udisks.
One final comment I would make is that are you certain the the mount options you’re trying to apply aren’t already defaults ? Check with the mount command after the drive is mounted what options are listed, also just because you don’t see an option doesn’t mean it isn’t there, for example I see that you have noatime in your list - noatime has been the default in Linux for many years now.
If you try to manually specify noatime in mount options and then check with the mount command after the drive is mounted, you will not see noatime listed - because noatime IS the default already. But if you were to specify atime (the opposite of noatime) you would see it listed when you check with mount.
I’m pretty certain that async and nodiratime are also defaults in linux for a long time now. (Running in sync mode is really slow) The only one I’m not sure about is big_writes.
Edit: Just realised that noatime and nodiratime are probably invalid for an ntfs file system in the first place, as ntfs doesn’t support the concept of an “access time” - that’s purely a unix file system concept, eg ufs,ext4 etc. So you shouldn’t be trying to use these options on ntfs in the first place.