Open Invention Network adds Microsoft's exFAT to Linux System Definition, Satan spotted throwing snowballs The Open Invention Network (OIN) has expanded the scope of its Linux System Definition to include the likes of exFAT and Android AOSP 10. Established back in 2005, OIN lays claim to being the world's largest patent non-aggression community, aimed at removing "patent friction in core open source technologies." It scored quite …
The 4GB limit on FAT32 dates back to 1980 when Tim Paterson was implementing FAT12 for QDOS/86-DOS while using some 4 byte block from the the Pile Control Block structure under CP/M for the file size under DOS. This was at a time when 8" DSDD floppies were good for 1.2MB and the relatively rare users of hard drives were content with 5 and 10MB capacities.
Ditto.
Mind you, with the amazing increase in disk capacity, the alleged default reserved space ought to be dropped to 1% anyway.
Can see why it's there, but I never exceed about 75% used on OS disks *, and storage disks don't need this at all. They should have a choice for neophytes to vary formatting according to intended use.
.
Actually once, but that was from trying out BTRFS, since when I have never strayed from the EXT4 true path.
To quote Wikipedia "Ext2 is still the filesystem of choice for flash-based storage media (such as SD cards and USB flash drives) because its lack of a journal increases performance and minimizes the number of writes, and flash devices have a limited number of write cycles. However, recent Linux kernels support a journal-less mode of ext4 which provides benefits not found with ext2, such as larger file and volume sizes."
You can turn off journalling when formatting:
To quote from my cheat-sheet:
Yes, you can - you can disable it at mkfs time with
mkfs.ext4 -O ^has_journal
or after you run mkfs.ext4, you can disable the journal with
# tune2fs -O ^has_journal
.
I would advise checking the syntax first...
ext2 has no journal. not sure you would want that. But I think ext4 has differences with ext3 that make it NOT compatible with ext2 for a few things. In any case, if I'm on FreeBSD and I attempt to use an ext2 built-in driver to read ext4, it fails. I need to use the appropriate FUSE driver for ext4. but the ext2 driver "works" with ext3, as long as you don't care about journaling.
so yeah depending on what you need, ext4 is a good thing, but not very compatible outside of the Linux world.
Yup, and it's not my first choice of filesystems, but we need an option that allows shared media access and still supports large files and volumes. Fat32 hasn't been able to cut if for a while in that regard.
I'd have preferred M$ open NTFS, but I'm not holding my breath on that, and ExFat addresses most of those concerns. Get it into the BSDs and we might be close to being able to support any modern media as portable storage under Win/Unix/OSX which is actually a fairly big deal when you start chucking projects and VMs around.
I still dream of universal Ext support without "magic" drivers, but I'm not holding my breath on that either, though there is no justifiable reason why it's not already baked into, well _everything_ by default.
"Get it into the BSDs and we might be close to being able to support any modern media as portable storage under Win/Unix/OSX which is actually a fairly big deal when you start chucking projects and VMs around."
UDF 2.01 is mature and supported in most - if not all - operating systems.
You were obviously downvoted by someone who doesn't know who UDF is.
The cross-platform solution has been staring everyone in the face for a decade or more, and while a decade ago some OSes in circulation had patchy support for UDF that is no longer the case.
Couldn't agree more - it's a pity the userland tools of many popular OSes neglect to support it well, so it takes a little bit of care (but is perfectly possible) to format a UDF disk in a way that works painlessly on every major platform. I can only guess Microsoft was too busy shoving ExFAT down everyone's throat via various standards forums (e.g. SD cards) to milk those sweet royalties, while UDF would have subverted that effort.
On the flip side it's nice to see that MS turned a corner on that particular issue, so if ExFAT does in fact see proper wide adoption without patent encumbrance then it's only going to be a win for everyone.
Yes, 256 G micro SD cards use it and should not be be reformatted without special tools.
Possibly most SDXC cards (64 G and larger?) use exFAT by default, I'm not sure.
Certainly actual gadgets have limits on format types and sizes, I think notable boundaries being 2G or 4G (4G can be SDHC or SDHC). Usually FAT
4G to 32G SDHC, always FAT32 by default.
64G to 2T SDXC usually exFAT. There is though no assurance that something that works with a 256 G SDXC card using exFAT will work with a larger card.
Most gadgets only work with cards up to a certain size and in the default format for that size. So while Ext2, 3 & 4 or NTFS will work on a laptop, they are unlikely to work on a camera, MP3 player, Zoom/Tascam recorder, portable video camera, phone or tablet.
I've not seen an SDUC card yet, which I presume also uses exFAT. Probably 2T to 128 T. I noticed 512 G cards yesterday and noted that the the 256 G micro SD card I bought early this year has dropped about £10 in price to about £28.
You need the reserved space for the onboard controller to silently swap dying memory locations.
occasionally I've needed to use exFAT to transfer very large files to a windows system on a USB drive, or if someone hands me a USB drive formatted from a windows system [which apparently defaults to exFAT unless you tell it NOT to, and that may not be very easy...]
In the past I've used FUSE. but as I use FreeBSD, is it part of OIN? Or are the people who make the FUSE driver part of it? In any case, I'll just use it anyway and ignore anyone peeking over my shoulder...
(I do not want my birthday taken away, no no, seriously, stop looking over my shoulder)
This post has been deleted by its author
This sort of thing is getting me worried these stories about a Linux kernel in Windows might be true. Worried because there's less likely to be enthusiasm for existing developers to do stuff for Microsoft, Linus gets pushed out and then Microsoft themselves start shoving dubious stuff in there. EEE accomplished.
Linus still owns the kernel, and he isn't going to allow that to happen. Neither will whoever takes over should he get hit by the proverbial bus. There will be an up-to-date and modern mainstream FOSS Linux kernel available that is not under the control of any corporate entity for far longer than you or I will be around to worry about it.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
