[Date Prev][Date Next][Thread Prev][Thread Next][Author Index][Date Index][Thread Index]

Re: [zzdev] Brainwave: Cell Contents

Dear Bek,

This shows how sharp you are.  This is just what
 Andrew and I have decided.

Keeping the contents outside the cells will make
 lots of things easier, including hand-editing and 
 eyeball viewing, fusion of datasets from different
 machines, etc.

More soon.  Also I think I sent out a memo on this


At 08:45 PM 9/25/98 +1000, you wrote:
>This occured to me while I was in the shower (why do brainwaves
>always come when you're in the shower and too wet and soapy to
>write them down (Hmmm, whiteboards in the shower, anybody?)).
>Ahem, anyway ...
>Is there any particular reason that all the data in a particular
>zigzag data file (We DO need a name for one of those.  Call it a
>'zag'?) (Then we can call the activity of using ZZ 'zigging' :) (but I
>digress ...) to be all in one "place"?
>Currently we store ALL the ZZ data for a single "database" in one
>file, usually zigzig.data in the current directory.  I recall people
>talking about interfacing ZZ with OSMIC and hence storing all the data
>in one OSMIC server.
>What about letting a cell contain a REFERENCE and then we can display
>the referent instead of the reference.  Like formulas in spreadsheets
>where you enter a formula but it displays to you the result, not the
>So if you entered:
>   http://xanadu.net/ZZ/
>as the only contents of a cell then ZZ will grab the page (this is easy
>in perl) and display THAT in the cell.
>Or if you entered:
>   osmic://servername/1-22,44-45,some-other-offsets
>it will query the OSMIC server on servername and construct the right
>display from the offsets you give it.  Etc.
>You could also, a-la Lynx, enter:
>   file://localhost/home/gossamer/notes
>and it'll grab the file off the disk and display THAT ... or
>   exec://localhost/file/name/of/a/program
>and it'll execute the program and display its OUTPUT!
>If you wanted to go the whole-hog you could make a referent the ONLY
>thing a cell can contain, and if the user tries to edit an empty cell
>you can invent a local file (say ~/.zigzag/some-random-name) and etid
>that.  Sorta like HyperWave invents filenames if they aren't
>   - You can edit the individual text files out of ZigZag if you want
>     to.  (This is something I'd like).
>   - Makes the data file smaller and easier to manage.
>   - You can get ZigZag to contain information that changes - eg I
>     could put http://www.slashdot.org/ in a cell and always see the
>     current slashdot headlines!  Or the weather, or the stock exchange
>     values or the MOTD or ... anything!
>   - There's Perl modules to support this sort of thing, so it's not
>     hard to code.
>   - You can no longer just give somebody a zigzag.data file and 
>     guarantee that's -all- the info they'll need to display the
>     dataset it contains.  But we could write a utiliy to run over
>     the cells and list all the relavent files fairly easily.
>   - Figuring out when to update the data from files that may change,
>     like URLs.  This could make startup slow if we have a zag that
>     contains 100 URLs and they all try to load at startup!  Need some
>     sort of local cache and some smart rules about when to update
>     things?  Possible though.  Also some sort of update and/or
>     write-lock for local files, same reason.
>   - Have to "know" which pages you can and can't edit and disallow
>     editing for certain cells.
>I keep thinking somebody else MUST have been through this before, is
>it a new idea at all?  But I like it a lot.
>Just noticed something ... I assume osmic://servername/offsets isn't a
>valid URL but can we register it somewhere/somehow so it is?  That'd
>mean any browser could display OSMIC files!!  Does OSMIC even have a
>standard port or anything?  I can alter the program trivially to work
>accross a network, if it doesn't already.
>: Gossamer     gossamer@xxxxxxxxxxxxxx  | Xanadoodler
>: http://www.tertius.net.au/~gossamer/  | And proud of it :)
>: This above all; to thine ownself be true, and it must follow, as the
>: night the day, thow canst not then be false to any man.
>: -- William Shakespeare, 'Hamlet'
Theodor Holm Nelson, Visiting Professor of Environmental Information
 Keio University, Shonan Fujisawa Campus, Fujisawa, Japan
 Home Fax from USA: 011-81-466-46-7368  (If in Japan, 0466-46-7368)
Professorial home page http://www.sfc.keio.ac.jp/~ted/ 
Permanent: Project Xanadu, 3020 Bridgeway #295, Sausalito CA 94965
 Tel. 415/ 331-4422, fax 415/332-0136  
PERMANENT E-MAIL: ted@xxxxxxxxxx
Quotation of the day, 98.09.25:
"Love is the marshmallow on top of the blood-and-guts sundae we call life."
 TN 98.07