workPage

Page and feed autogeneration

Users should have certain pages autogenerated in a directory named with their username.

Various feed seem very useful, such as contextual navigation feeds (tables of contents), image slideshows, and indices.

Here is some further research on the subject of how to autogenerate useful tools and pages:

From: Michael Hoffman (michael@pdr-is.com)
Subject: Every site needs Contents, Index pages -- CGI!?
Newsgroups: alt.hypertext, comp.infosystems.www.authoring.misc, comp.infosystems.www.authoring.html
Date: 1996/03/11

Does anyone know of CGI tools to automatically generate well-formatted
Table of Contents and Index pages? I might even write them myself.
I'm really into organizing information. There are probably some CGIs
out there that I could modify. For all those sites out there that
should have these navigation tools, I would like to be able to point a
spider at the site and generate a TOC and Index page to find my way
around and get a good handle on what's really there. The spider could
also transmit these two pages to the site's webmaster and encourage
them to include the pages so people can find their way around the
site. People will be almost forced to do this as web sites become
*much* larger than they currently are. Everyone's site is going to
become like a substantial book. The current navigation tools are not
nearly organized enough.

The power of the detailed outline is unsurpassed. And hypertext in no
way obviates or supercedes the traditional index.

NAVIGATION TOOLS FOR WEB SITES

Even the "well-designed" sites are very badly designed as far as
retrieving information and finding your way around. These sites are
elegantly mute about what they contain. 5 pretty pictures on the home
page does not constitute a navigation tool. This mysterious challenge
is really easy to solve, as follows.

The Need for Better Site Navigation Tools - in fact, traditional Table
of Contents and Index!

Every site should have a complete Table of Contents and Index. No one
would be lost in hyperspace. This would require CGI tools to
automatically generate the TOC and Index. The "keyword search" tools
can be used but people should actually *show* the Index document,
because visually browsing the index is much better than having to
blindly enter various keyword search terms. The CGI should generate
well-formatted Contents and Index html pages. People like first
seeing a simple home page with a few main links. The next page
offered should be the complete Table of Contents. At the bottom of
each page should be a 'search' and/or 'index' link, taking you to a
true, browsable Index page.

The Web information design standards currently have the same problems
as Microsoft Windows Help used to have -- confusion over the meaning
of 'outline', 'contents', 'index', and 'search'. Most "search" boxes
are not full-text search -- they actually only search the "keyword
index" page, which is not itself shown. The Keyword list is built by
automated guesswork and looking for defined keywords, which no one
uses. No wonder the Search engines so often return no results. The
web standards for retrieval paths are very poor, almost nonexistent,
compared to the traditional Table of Contents and Index.

Some sites take a feeble step in this direction with an "outline"
page... but instead of the 5 icons on the home page, they list about
every fourth document -- this is equivalent to a table of contents
that shows levels 1 and 2 but not 3, 4, or 5.

And

From: Cliff Urr (pp000305@interramp.com)
Subject: Re: Every site needs Contents, Index pages -- CGI!?
Newsgroups: alt.hypertext, comp.infosystems.www.authoring.misc, comp.infosystems.www.authoring.html
Date: 1996/03/12

I think for little itty-bitty sites, brute-force string indexing is
fine. But for any site that is substantial, the brute force approach
is lousy. Hand-made indexing, so to speak, shows relationships and
levels within and between distinct types of information that can be
very, very helpful in enabling quick acccess to the right place. The
brute force approach also does nothing with regard to omissions. If I
put in "cars" at a site that has "autos" instead, I will get no search
results, which is very misleading and wrong. Good indexing can include
thesaurus structures, such as "See" or "See also" to disambiguate
terms or concepts, to use a computer nerd phrase. (Try using
brute-force indexing in the Index Medicus database - I guarntee you
will get dreadful search results in most instances.)

Cliff Urr
James Martin & Co. (Library)
Fairfax, VA
/>cliff_urr@interramp.com

kflood@iesun2.ie.oracle.com (Kevin Flood) wrote:

>Regarding indexes, they are redundant on the web. If a site implements a
>proper search method, all of the words in every document become virtual
>index entries.
>Regards,
>Kevin.
>--
>==============================================================================
>Kevin Flood European Porting Centre, Email: kflood@ie.oracle.com
>Technical Writer Maretimo Court, Phone: +353-1-2834700 (X 248)
>Oracle Corporation Temple Road, FAX: +353-1-2834732
>.................. Blackrock, ..............................
>.................. Co. Dublin. ..............................

And

From: Michael Hoffman (michael@pdr-is.com)
Subject: Re: Every site needs Contents, Index pages -- CGI!?
Newsgroups: alt.hypertext, comp.infosystems.www.authoring.misc, comp.infosystems.www.authoring.html
Date: 1996/03/19
[snippage]

Windows Help was confused, at first, about the meaning of "Contents",
"Search", and "Index". But finally WinHelp has evolved fully. And
you know what it looks like? A book. This suggests that a fully
mature hypertext design comes back around to the book paradigm for
information structuring and access. The traditional book contains
these 2 info-access devices: Contents, and Index. Online is
well-suited to the Search type of info-access mechanism. Ideally, you
must have all 3. Oh, books and webs also have this lesser tool:
Overview.

Typical web sites today have:

o No Table of Contents
o Overview
o No Search
o No Index

There is *no reason* why hypertext must be associated with "lost in
hyperspace". With navigation maps such as a traditional hierarchical
outline -- a good table of contents with clear, dry headings -- you
should always know where you are.

Give me a rat's nest of topics, and I will produce an orderly outline.
Hierarchy is good. Order is good. People try to use hypertext to
liberate themselves from the oppression of order. They say that
hypertext "is" an exploration space, like Myst. But what about
usability! If you care about usability, then provide a good,
detailed, well-worded table of contents. You can have a pretty home
page, with its overview of main areas of your web. But the first link
on the home page should be Site Map or Table of Contents. In the
navigation bar, the first item should be Home. The very next item
should be Map or "Contents".

Let me define these poorly understood terms. They should be totally
obvious, but even Microsoft completely botched their terminology for
WinHelp 3.x. People think that hypertext is so absolutely
unprecedented, they can throw out all the conventions of terminology.

o Table of Contents: This appears at the front or top of a
"document". Should include H1 *and* H2, H3, possibly H4 and H5, for
*every* page or section in the document, book, or web site. It is
*hierarchical* and it is *long* and *complete*. Not just 5 or 10
items! For that, see "Overview".

o Overview: Lists only about 10 entries. Example: home page, or
survey of a book placed in the front matter before chapter 1.

o Document: This comprises multiple topics, sections, or pages. For
example, a medium-small web site, a book, or a long article.

o Library (Collection, Bookshelf): This comprises multiple documents.
For example, a large web site.

o Search: Full-text search, able to find a specific multi-word string
anywhere in the book or web site.

o Index: Occurs at the end or bottom of a document. To build an
index, go to each topic, page, or section of a document, and think of
3 terms by which a user would try to look up that topic.

o Master Index: Helps you find what document (book or web site) to
search in. Broader and shallower than any one document's index.

Edit  |  workFlow  |  Subscribe
Language: fr  | it  | de  | es  | pt  | ar  | he  | da  | nl  | zh  | ja  | ko  | none 
Author: dtd
skyCalendar

This Version:
Archived at: https://www.skybuilders.com/Users/Derek/Spider.20020716122305.html

Requests
 Version: 2908 | Series: 2908 

Search: Site | Web | Groups