From mboxrd@z Thu Jan 1 00:00:00 1970 From: Hans Reiser Subject: Re: Is it good for a PDA fs? Date: Sun, 23 Jun 2002 12:17:48 +0400 Message-ID: <3D15842C.1080100@namesys.com> References: <20020622115207.E23457@gerf.org> <20020622175746.I13075@parcelfarce.linux.theplanet.co.uk> <3D14B8E9.4020808@namesys.com> <20020622230748.A1981@namesys.com> <3D14E47E.5080300@namesys.com> <20020623105057.A5090@namesys.com> <3D15762A.3000507@namesys.com> <20020623114536.A1003@namesys.com> Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Return-path: list-help: list-unsubscribe: list-post: List-Id: Content-Type: text/plain; charset="us-ascii"; format="flowed" To: Oleg Drokin Cc: Chris Dukes , The Doctor What , reiserfs-list@namesys.com Oleg Drokin wrote: >Hello! > >On Sun, Jun 23, 2002 at 11:18:02AM +0400, Hans Reiser wrote: > > > >>>>I think I probably care more about downloading and trying the latest >>>>kalendar than installing a new kernel on it.;-) It is nice to have ssh >>>> >>>> >>>Yes, exactly. I do not see how you (or any other user) would benefit from >>>using anything but vfat in CF, but disadvantages are pretty clear. >>> >>> >>Space would be saved if the journal was shrunk in size.... >> >> > >VFAT (fat32) addresses 512 byte blocks while reiserfs now addresses >4096 byte blocks. (not this can be theoretically decreased to 1024 byte blocks >with a patch, but in this case tree will grow very quickly and will reach >height of 5 way too soon). And while on reiserfs tails are not packed >for files over 16K (and not packed for smaller files in some cases to to avoid >"seeks"), it is impossible to spend more than 511 bytes of unused space >on vfat no matter what file size is. (vs max of 4095 bytes wasted on >reiserfs in unlucky case). >So right now I do not see noticeable space savings. > End of file space wastage is critical for small files not large ones. For small files we do much better than VFAT. I don't know what overhead they have for various kinds of metadata though..... It is probably true that VFAT does a better job for CF than for usual hard drives though. >Of course I have not performed precise measurements so this is only >a theory. > >Bye, > Oleg > > > > -- Hans