[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Gnu-arch-users] FOSDEM substitute
From: |
Thomas Lord |
Subject: |
[Gnu-arch-users] FOSDEM substitute |
Date: |
Thu, 23 Feb 2006 12:08:08 -0800 |
Since I won't be at FOSDEM I'm not preparing a talk or slides. I am,
however, writing a brief series of short essays about the topic I
meant to speak about.
This is the first in that series of essays.
The Kings English: What Do Links Point At?
Consider a typical W3C, XHTML link:
<a href="http://google.com" [...]>google</a>
By the time your web browser fills in form details, cookies, and so
forth a link boils down to just:
~ a URL
~ an HTTP request
To "follow the link" is to send that request to the host indicated
by that URL and receive a reply.
What happens when you issue a request? One way to look at it is
that you are calling into a very large, distributed program with
many well known subroutines and data tables. For example, this
program looks up the host name in the DNS tables ultimately
maintained by ICANN; it looks up the IP address in routing tables.
When execution reaches the final server, the operating system calls
Apache which in turn calls your favorite module. There is a large
hierarchy of subroutine calls, right down to subroutines responsible
for just a tiny part of the page returned to you (such as an ad or
the content-space on your blog page).
People modify that very large program in real-time. The rights to
make modifications to some subroutines and tables in the hierarchy
are traded, loaned, and leased. For example, a registrar is in some
sense responsible for all of the code to handle an entire top-level
domain. A domain "owner", for the duration of their lease, is in
charge of the subroutines for just that domain. The rights to
modify the big distributed "web program" are recursively carved up
and leased.
A link indicates a URL plus a request. In other words, a link is
a pointer to a piece of commercial real estate -- a pointer to
location in the subroutine tree. A link is a short-hand name for
specific legal rights that somebody, somewhere, may rent.
You may think to yourself "I will link from my essay to the
Constitution," or "I will link from my essay to the Wikipedia
article on Thomas Jefferson". In fact, though, you can do neither.
Instead, the best you can do is to refer to those documents
indirectly. You can say "I link to the document stored in room A,"
or "I link to the document stored in room B". You can only hope
that those rooms will continue to contain the document you meant for
as long as people want to read your essay.
Hyperlinks are some new kind of word -- a new vocabulary. Sometimes
they are used as nouns, verbs, adjectives, interrogatives ... but in
all cases they have a specific meaning -- a definite thing referred
to.
With the language of W3C hyperlinks, we may speak only of property.
* Positive Consequences
There can be no doubt that there are positive consequences of having
to speak only in the "King's English" of W3C links:
The "right to define a subroutine" of the big web program is often
best exercised by building. In an effort to improve their web
properties and put them to productive use people have poured
billions if not trillions of dollars into networks, machines,
software, and services. The rent charges for web real estate have
driven a privately-funded building-out of the net in general.
The syntax and psychology of domain names is such that some names
naturally have higher value than others. `coca-cola.com' is likely
to be worth more than `joes-aunt-betty-m-in-florida.net'. Some
domain names -- simply by virtue of the name -- have a greater
chance of being put to greater productive value. The stakes are so
high that millions are poured into the most valuable properties
(e.g., `ibm.com') leading to a trickle down of investment that
benefits the web as a whole.
Above all, everything on the Web is paid for, one way or another,
with very little of that funding coming from charity or government
spending. Significant portions of the web territory are productive
enough to return a profit.
All of that nice economic consequence is the direct effect of
limiting links to point only to real-estate. It is valuable
for many, especially corporations, to put stable, linkable content
at an easy to find place on the web. The only way to do that
is to own and improve virtual web real estate. The flurry of
build-out spending naturally follows.
* Negative Consequences
We could define a "Web Author" to be:
A person who is able to create stably linkable
documents on the web.
We could define an "Archive Quality Hypertext" to be:
A collection of mutually linked web documents
whose content can be freely copied and served
from arbitrary locations without loss of semantic
content.
The link-only-to-real-estate property of W3C links (as commonly
used) means that the only way to be a Web Author is to hold the
lease for part of the URL + Request real-estate and then to exercise
your rights under that lease to modulate the contents of that
real-estate. The right to write, on the web, must be purchased and
is quite expensive. (The right to write on the web is *so* precious
that very tiny slivers of it -- such as a blog account -- are
valuable enough to be given away as promotional items.)
The link-only-to-real-estate property of W3C links also means that,
for the most part, archive quality hypertexts simply *do not exist*
on today's web. If four biotech labs around the company each put up
wikis with lots of linking between them, odds are not good that the
content after five years can be copied cleanly and set aside in
pristine form in the archives of the Weidner Library.
* The Challenge, Then
Can we improve our implementation of "hyperlink" in such a way as to
move some items out of the "negative" column and add some to the
"positive" column? Can we do this without unwanted, unintended
consequences?
-t
- [Gnu-arch-users] FOSDEM substitute,
Thomas Lord <=