[Date Prev][Date Next][Thread Prev][Thread Next][Author Index][Date Index][Thread Index]
Re: [zzdev] :zz: "OS theology
- To: zzdev@xxxxxxxxxx
- Subject: Re: [zzdev] :zz: "OS theology
- From: Andrew Pam <xanni@xxxxxxxxxxxxxxxxx>
- Date: Fri, 2 Oct 1998 04:08:02 +1000
- In-reply-to: <3.0.3.32.19981002020340.008943c0@xxxxxxxxxxxxxxxxxxx>; from Ted Nelson on Fri, Oct 02, 1998 at 02:03:40AM +0900
- References: <3.0.3.32.19981002005224.00888850@xxxxxxxxxxxxxxxxxxx> <3.0.3.32.19980630000324.014597e0@xxxxxxxxxxxxxxxxxxx> <3.0.3.32.19980630000324.014597e0@xxxxxxxxxxxxxxxxxxx> <19981002014124.O2218@xxxxxxxxxxxxxxxxx> <3.0.3.32.19981002005224.00888850@xxxxxxxxxxxxxxxxxxx> <19981002020701.R2218@xxxxxxxxxxxxxxxxx> <3.0.3.32.19981002020340.008943c0@xxxxxxxxxxxxxxxxxxx>
- Reply-to: zzdev@xxxxxxxxxx
On Fri, Oct 02, 1998 at 02:03:40AM +0900, Ted Nelson wrote:
> >I can't presently see any advantages to developing hardware device
> >drivers using the Zigzag structures, and considerable disadvantages.
> >Feel free to convince me otherwise with examples.
>
> Hardly. Let it stand as a religious statement:
> "Files (named lumps in fixed places, combining
> various types of data), and hierarchical directories,
> are grotesque and stupid structures whose
> grotesquerie is not recognized because they are
> so familiar.
> "Similarly, program and data references based
> on string-scans of the names of such files and
> directories, and on paths among them-- the
> fundamental underlying referential structure of
> all operating systems today-- is a mistake.
> "For discrete elements now packed into files,
> as well as discrete references currently made
> to files, directories, inodes, processes and whatnot,
> cellular reference could be better;
> "ZZpaths provide a far more general referential
> structure than pathnames.
> "The general alternative here is a world where
> plexes of smaller discrete elements, connected through
> ZZpaths, replace 'files' to a large extent, and where
> ZZpaths become a more general referential structure
> to replace OS pathnames."
I fully agree with you. However, you are still addressing the
issue of structuring, filing, naming and retrieving of information,
all of which Zigzag is well suited to. The point I was making was
about the unsuitability of Zigzag to low-level hardware control, a
completely different issue. Files, directories, paths, inodes etc. are
all filesystem concepts, and may be used as an interface to devices; of
course we could replace this with a cell visualisation of (and interface
to) devices, which would still be provided by an operating system layer
which I believe is outside the direct scope of Zigzag.
Cheers,
*** Xanni ***
--
mailto:xanni@xxxxxxxxxx Andrew Pam
http://www.xanadu.com.au/ Technical VP, Xanadu
http://www.glasswings.com.au/ Technical Editor, Glass Wings
http://www.sericyb.com.au/sc/ Manager, Serious Cybernetics
P.O. Box 26, East Melbourne VIC 8002 Australia Phone +61 3 96511511