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



I've recently used Google to search for "GZigZag," and here are some
results that caught my attention. I think it is interesting and useful
to know how some people understand (sometimes: misunderstand) and judge
ZigZag. The snippets are in no particular order.

- Benja

Joshua Allen <http://static.userland.com/userLandDiscussArchive/msg021373.html>:
     For point 1, I just wanted to make sure that the impact
     of this sentiment in OSS is not dismissed. One example I go to
     is Ted Nelson's "ZigZag" stuff (gzigzag @ sourceforge.net).
     He has been working on this for a few years, but still
     most real people have not seen it or benefitted from it.
     The problem is, the only developers he can get working for the
     project insist on developing for Linux and Java, and Windows
     is an afterthought. This tool has huge potential to "wake
     up" the web, but has gone almost nowhere. If you contrast
     napster, which ran on Windows, to many OSS projects that
     arbitrarily and capriciously decide that Windows isn't clear,
     you can see how I feel that an anti-MS sentiment can often
     greatly reduce the overall impact of an OSS project. This is
     just one thing I see as a recent side-effect of the mood; I am
     sure that others can think of other good or bad things that
     come from this attitude. So I was just saying, "it's true, and
     has interesting consequences, so please don't dismiss it too

Sanemagazine horoscopes <http://www.sanemagazine.com/horoscopes183.html>:
    [Horoscopes. GZigZag!. Ted Nelson's ZigZag given a 
     Java implementation.] 

(Benja: No, I have no clue why it's on a horoscope page, either.)

Usability Forum <http://www.usability-forum.com/veranstaltungen/archiv/25May2001095437.shtml>:
    Cfp: Zigzag Workshop, Hypertext 2001
    25 May 2001

(Benja: What stuck me about this is that the workshop would be mentioned
as a "usability forum." ZZ certainly has the potential to become very
usable, but the current implementations... well... ;o) )

Will Newton <http://www.ctanet.fr/~sheflug/mailarchive/2001/02/msg00152.html>:
    > I'm intrigued by the idea of 2-way links but I
    > don't have any idea what Xanadu is or how it achieves 
    > them.  Hypercard has/d only 1-way links I think.

    Xanadu (or some version of it) is available from 
    somewhere. :) www.xanadu.org is offline, ah, 
    www.gzigzag.org is where it's at. I think
    Xanadu disappeared and is now zig zag, or something. :)

Kevin Turner <http://www.twistedmatrix.com/pipermail/twisted-python/2001-July/000102.html>:
    >       Example 2:
    >         Information management.  

    Another good place to look for examples of information
    management (in addition to what churchr mentioned) is
    at Zigzag, Ted Nelson's brainchild.  There's an
    implementation now that's at least far along enough
    that you can play with it at http://gzigzag.sourceforge.net/ .
    I don't feel it's quite to the point where it's ready to take
    off on its own, as the applications for it are rather
    incomplete and the only chunks of data you can really use in
    the default interface are one-line strings, but it's a toy from
    which you might be able to learn something about dense
    bidirectoinal hyperlinking.

NetworkWorldFusion Compendium <http://www.nwfusion.com/columnists/2001/0326compendium.html>:
    Making your data go zig-zag

    You remember Ted Nelson. He's the guy who first began
    working on a hypertext system way back in 1960 - and
    who is still working on it. Now he's taken a wee detour
    with a new project called ZigZag that is sort of a
    cross between HTML and object-oriented programming. At
    least, as far as I can tell.

    The basic idea behind Zig Zag is to have discrete bits
    of information that can be recombined in an infinite
    number of ways (the data is stored in "cells," which are
    connected via "dimensions") to create new documents,
    lists, whatever. Think of your family geneaology program
    (what, you don't have one?) and how useful it would be to
    quickly connect all those lines.

    Gzigzag is an open-source implementation, if you want to
    play around with something that is: 

          locally rational, globally paradoxical, yet
          somehow comprehensible. 

Jack Park <http://www.infoloom.com/pipermail/topicmapmail/2000q3/000054.html>:

       Just to toss out an implementation of a multidimensional
       database, consider GZigZag.  It may not mean all that
       much to topic maps, but it may spawn some thinking about
       user interfaces and so forth.
       http://gzigzag.sourceforge.net/ <http://gzigzag.sourceforge.net/> 
       Be sure to read the faq.
       See also http://www.xanadu.net/zigzag/
       "ZigZag is a delightfully simple way of operating
       structures. As such, it has its own uses simply as a
       personal information manager for people who
       like multidimensionality."
       You can download source and a jar file.  The jar file
       is a hoot to run.  I have no idea how to use it yet, but
       it sure is pretty. Documents, including a user's manual
       are at http://gzigzag.sourceforge.net/docs.html

Flemming Funch <http://groups.yahoo.com/group/ourownthoughts/message/287>:
    In browsing Steve Danic's memes.net site I noticed a note 
    on a new  implementation of Tel Nelson's (xanadu.com) ZigZag
    program, which I'm very interested in. It is an open source
    project hosted at 

    ZigZag is very hard to wrap one's mind around, and it isn't
    implemented in any remotely user friendly way, but it has
    some of the essentials that I'm looking for in a system for
    organizing thoughts and information. It is multi-dimensional,
    with an arbitrary number of dimensions, and you have travel
    through it along any 2 or 3 dimensions at the same time. It is sort 
    of a multi-dimensional matrix cube that you can place data items in.

    It might be an idea to approach the people working on it to
    let them know of this group.

Stephen Danic <http://groups.yahoo.com/group/ourownthoughts/message/290>:
    > It might be an idea to approach the people working on
    > [gZigZag] to let them know of this group.

    I have briefly corresponded with Tuomos Lukka, the project
    maintainer for gZigZag. I mentioned my interest in starting
    a knowledge apps foundry on sourceforge. I didn't mention
    OurOwnThoughts or Minciu Sodas though.

Flemming Funch <http://groups.yahoo.com/group/ourownthoughts/message/298>:
    > > most things I want to remember belong
    > > in a bunch of folders at the same time, and I can't 
    > > remember which of those folders I actually put it in.

    > I.e., you need something with the possibility of multiple
    > inheritance, multiple tags (sets) assigned to your information
    > items, and corresponding convenient browsing possibility.

    I think so. And I'd love it to be some inherently simple
    paradigm, so that multiple tags on an item simply constitute a
    few more dimensions to what is stored. Multiple inheritance -
    I suppose so - except for that I'm not sure a traditional
    object-oriented metaphor quite covers it. Which might be my 
    lack of understanding, of course.

  << Benja's snip >>
    > You can download source and a jar file.  The jar file is a
    > hoot to run.  I have no idea how to use it yet, but it sure
    > is pretty. Documents, including a user's manual are at
    > http://gzigzag.sourceforge.net/docs.html 
    > <http://gzigzag.sourceforge.net/docs.html>

    Can anybody tell me in a couple of sentences how to run
    jar files in Windows2000 or on a Mac? Java remains a
    mysterious closed world to me.

Eric Armstrong <http://www.bootstrap.org/dkr/discussion/2337.html>:
    Jack Park wrote: 
    > Of primary interest to an OHS is the idea that an
    > information cell need not be cloned into several places
    > for different views. Rather, one cell can be a member
    > of an infinite number of 'threaded' views. That way,
    > only one cell ever exists for a particular concept, and
    > that lone cell, once modified, is immediately available
    > to all views which would use it. 

    This is the basic idea of multiple-parenting. The core 
    notion is that structure is divorced for content. A
    structure consists of a set of links

  << Benja's snip >>

    Note that simple node-reuse is relatively easy. 
    It sounds as though the "cells and dimensions" 
    idea may be focused on that track (although I'm 
    not entirely sure that is the case). 

    The ability to reuse structures is quite a bit 
    harder, but of great importantance, imho. 

    The way I've seen to reuse structures is 
    to put the root of the subtree in a "wrapper", 
    that identifies it as different from "native" 
    nodes in the tree. That makes it possible to 
    add type information, to identify the kind of 
    information that is being included. That's 
    important so you can hide elements by type. 

Eric Armstrong <http://www.bootstrap.org/dkr/discussion/1689.html>:
    Bill Bearden wrote: 
    > Try the following. 
    > http://www.sfc.keio.ac.jp/~ted/zigzag/xybrap.html 
    Thanks. I now have a clear idea what he's talking about. 

    I'd love to see a report on user's experiences 
    with that interface. I suspect the results will be 
    rather negative. (The author will thereupon blame 
    the users, if history is any guide.) 

    From my quick inspection, it looked as though the 
    underlying is model is simply one of multi-list 
    trees, with links. (A multi-list tree is one in 
    which node has multiple sets (lists) of children. 

    Those are extremely powerful model for data 
    management, especially when combined with 
    internal links. The problem has always been how 
    to present the information to the user. 

    His solution is to use the concept of dimensions, 
    which is reasonable. Each sublist is, in effect, 
    a different vector. But the fact that dimensions 
    don't appear until you visit a node, and that 
    the context (tree in which the node appears) is 
    lost when you visit a node (judging by my quick 
    scan) severely hampers the usability of the system. 

    My thought has always been to portray each dimensions 
    as a column in a table. Each node in the tree therefore 
    has a table under it, with one column for every 
    element vector under it. The "special case" of that 
    presentation is a standard outline, where each node 
    has only a single sublist (vector of subelements). 

    Using the multi-column model allows you, at least in 
    principle, to visit any given vector while retaining 
    as much of the surrounding context as you desire. In 
    practice, of course, we are limited by the woefully 
    inadquate quantity of screen real estate at our 
    disposal. I expect LCD research to solve those problems 
    over the next 10-20 years, though, so prospects for 
    such a view seem very interesting. 

    Admittedly, these notes are based on a very quick skim 
    of the article, so I may have misconstrued his real 
    meaning. By I have hard time listening to that much 
    belly-aching, so it is difficult to peruse his material 
    in any depth. 

Grant Bowman <http://www.bootstrap.org/lists/ohs-dev/0587.html>:
    ZigZag has a ZigZag Transfer Protocol (ZTP).
    I wonder if this compares favorably with Corba,
    WebDAV, OGS or others. 

(Benja: Grant's forwarded a GZZ release message, saying that GZZ now has
preliminary ZTP support.)