Date: Thu, 17 Apr 1997 12:00:31 -0700 (MST) From: Terry Lambert <terry@lambert.org> To: thorpej@nas.nasa.gov Cc: davem@jenolan.rutgers.edu, bsdhack@shadows.aeon.net, spidaman@well.com, hackers@freebsd.org Subject: Re: Feasibility of porting Linux filesystem code? Message-ID: <199704171900.MAA00341@phaeton.artisoft.com> In-Reply-To: <199704171706.KAA17594@lestat.nas.nasa.gov> from "Jason Thorpe" at Apr 17, 97 10:06:51 am
next in thread | previous in thread | raw e-mail | index | archive | help
> So, you read the papers that have been published on it, and implement > something that does basically the same thing. You only have to > "reverse engineer it completely" if you want them to be compatible > (i.e. want to be able to plug a disk from your SGI into your PC or > whatever). Personally, I don't care about that too much. They would have a hard time enforcing against even a binary compatible implementation. Look at the NTFS stuff that Microsoft can't enforce (there is a read-only driver that loads as a DOS TSR; it's available from the O'Reilly WWW site). Also, there's: http://www.sgi.com/Support/DevProg/Forum/forum96/proceeds/Filesystem_for_IRIX_6_2_and_Beyond/overview.html The claim thair HSM module has been submitted to X/Open as a proposed standard. They appear to use DOW to delay block allocation. They appear to recover at mount time (I am suspicious of this type of thing because it can't reasonably protect against hardware failure like an offline tool might). They export the transaction log facilities to the FS consumer. Basically it's sufficiently documented that I doubt you could step on any toes. Regards, Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199704171900.MAA00341>