> With regard to modern filesystems, the intrinsic limitations that you > and the previous poster pointed out seem to rule out the LEAN > filesystem as well. Am I right in my inference?
LEAN's internal structures are designed to be simple, so writing a (slow) driver that uses little memory to access it should be possible. However, it doesn't store/honour short name aliases either, so the long name difficulty is essentially the same. The lack of DOS interface support (eg for extended attributes) is also the same, though LEAN FS itself does potentially provide some permissions and ownership numbers, more precise timestamps, and some form of extended attributes. (See the current specification for details.) > It's strange, because the LEAN filesystem was designed to work on > FreeDOS-32, which is a DOS too, judging by the name. Can someone > explain the incongruence. Yeah, that might be a bit misleading. The FreeDOS-32 kernel runs in protected mode and is an entirely different thing from the FreeDOS kernel (which runs in V86 or real-address mode). Additionally, DOS (SFN) compatibility was not yet considered in the LEAN FS design - though it is still being extended from time to time. (Storing SFN aliases within an extension to the current directory entry format would be trivial, but all drivers should then be required to honour such existing aliases (even if they don't create aliases themselves) - to avoid creating a native/long name exactly matching that alias.) > BTW, is FreeDOS-32 dead or what? It hasn't seen a lot of development in recent years. The most recent major update (0.6.0) to the LEAN FS specification itself was released on 2010-03-22, with the most recent minor update (0.6.1) on 2011-04-09. No idea when the actual FreeDOS-32 code was last touched. > What I was would be willing to settle for, eventually, is ANY > filesystem with extended attributes that can be used from within > (Free)DOS. But from what you have told me the prospects do not seem > very bright right now... :( Well, you could design the required interface extensions and then implement the drivers to conform to those =) Other than that, yeah, seems unlikely. Regards, Chris ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ _______________________________________________ Freedos-devel mailing list Freedos-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/freedos-devel
