Thoughts on FAT64/exFAT

Yeah, yeah. M$ has released the next generation filesystem for large capacity flash media (iow, the portable USB sticks/external hard disks) and claims to offer “seamless interoperability”. Well, lets sit down and dissect.

It brings back the memories of struggling to mount a NTFS partition my friends had, and smiling a futile attempt to tell them that “Ugh, NTFS isn’t supported yet”. It was not long back. We got NTFS readonly support later, and I remember searching and installing the Kernel module to mount partitions of the dual-booting machines in our lab.

We got reliable NTFS write support with ntfs-3g much later, IIRC by 2006.

And the story repeats.

But no, there’s already an experimental driver for exFAT, thanks to OGAWA Hirofumi. It is reported to work well, but there are issues. Still, he had to hack the image created by Vista to create that driver.

Will there be a “non-sign-your-life-away spec” open to the developers, ye preachers of interoperability?

Advertisement

Squashfs in mainline kernel 2.6.29

One thing I loved following the Linux Kernel development discussion is the wit of those “mutant super geniuses” and humorous dialogues exchanged between them.

Though Squashfs has been in use for a long time – used by virtually every distro with a LiveCD, it wasn’t part of the mainline Linux kernel. The merge request came during the 2.6.29 merge window. It was found to be a ‘not much invasive’ code during the review, and finally pulled by Linus.

And, look at the comments:

• “Merge it! We have long needed a filesystem named after a vegetable” – Andrew Morton.
• “You make a very compelling case.” – Linus Torvalds.
• “And vegetables are much healthier than FAT…” – Alan Cox.
• “I always preferred my squash with butter, it appears 2.6.30 will be a very tasty kernel.” – Harvey Harrison.

😉