|
|
 |
|
|
Do NOT use your browser's back button. Use the Inbox button to refresh your messages.
|
|
| Message
1 of 7 |
| From: |
"Bob Hopgood" <bhopgood@brookes.ac.uk> |
| To: |
"'Kevin McCurley'" <mccurley@almaden.ibm.com>
| | CC: | <bhopgood@brookes.ac.uk> |
| Date: |
Wed, 5 Mar 2003 12:41:49 -0000 |
| Subject: |
RE: WWW2003 final submission guidelines fail to follow the W3C recommendations. |
|
Kevin,
Not sure who should be replying (probably not me
but here goes). IW3C2 is interested eventually in getting all the papers for all
the Web Conferences online with good linkage between them. It is quite a big job
as the early Conferences have a range of styles and formats. Wendy Hall's Group
at Southampton have a system that can automatically scrape the Title, Author
Affiliation, Keywords, References etc information from the way most of the
papers are defined and put this into a hypertext database without too
much human intervention. That is pretty well complete apart from one or two
years. It also is able to check and automatically link to Paper
References even when the link is not given.
We are
aiming at getting a format that everybody can agree to for the Conference in New
York. So we were trying to do a half way house for Budapest where ACM, Wendy and
the Conference people who were doing the online version would be happy. The
styling is nearly a copy of what ACM requires and uses their rules. ACM may
produce the CD version and want an electronic version so the electronic
version is not just a concern of IW3C2. (Probably worth making the point that
IW3C2 is not equal to W3C. One runs the Conference and the other generates the
standards).
I
don't have any real disagreement with the points you make but the solution for
2003 was a pragmatic one based on where we were. It has been a hard job
convincing people that it was now viable to expect authors to submit papers in
XHTML format even.
The
multiple authors with the same affiliation is an ACM problem. That is what they
state they require. One of the things we were trying not to do was to get people
to put the papers in using two different formats that would double the work they
had to do.
I will
make sure you get a chance to comment on the format proposed for New
York.
Bob
I am
writing to all of you regarding the submission guidelines for the WWW 2003
conference. As I was preparing my final version according to the submission guidelines, it became apparent to me that they are inappropriate
for a W3C conference.
The
acceptance letter for this conference says that both pdf and XHTML should be
submitted, which is completely reasonable. Unfortunately the guidlines
for submitting XHTML are misguided. The problem is that METADATA IS NOT
THE SAME AS VISUAL MARKUP, and the two should not be mixed. The
guidelines specifically state that metadata should be enclosed in h1, h2, and
h3 tags with class attributes to indicate the type (e.g., author and
affiliation). This is simply the wrong way to encode metadata into a
document.
According the the W3C
standards, there are at least two ways to include metadata according to the
XHTML specification.
- use the <meta> tag of the
HTML 4.01 specification (see section 7.4.4 of the spec
http://www.w3.org/TR/html4/struct/global.html#h-7.4.4 )
- the use of RDF in an external
namespace as described in the XHTML specification (see section 3.1 in
http://www.w3.org/TR/xhtml1/ )
Neither of these are recommended in the submission guidelines at
http://www2003.org/www2003-submission.htm
To see an example of why the proposed method is a bad idea and why the
XHTML guidelines separate metadata from visual layout, you need only consider
the paper that I am working on. This paper has six authors with a single
shared affiliation and address. The natural way to lay this out visually
would be to list the authors, followed by their shared affiliation and address. The ACM has their own weird rules about how to visually place
six authors on their proceedings, and each publisher will have their own rules
(some require affiliations in footnotes). Unfortunately, the guidelines
say that an H2 tag with affiliation should accompany every H1 tag specifying
the author. This mixes the metadata binding of the affiliation to the
author with the visual layout of the affiliation with the author, which
clearly does not make sense in this circumstance. If I follow the
guidelines, then I end up with something that looks visually comical, with the
same affiliation and address repeated six times. If I use good
typographical practices, then I fail to encode the metadata to indicate the
affiliation and address of the authors. This one example of why it is not a
good idea to mix metadata and visual layout.
ACM's digital library requests PDF submission formats,
so I presume that the XHTML version is requested specifically by the W3C.
The confusion between visual layout requirements and structural
requirements for machine-readable metadata is a long-standing problem in the
web, and is clearly one that the W3C has been trying to address with the
semantic web initiative. If the W3C is going to have these well thought
out standards for metadata and structure on the web, then it seems to me that
authors for the premier web conference should at least be encouraged to use
them correctly.
Kevin
McCurley
IBM Almaden Research
Center
Phone:
+1-408-927-1838
email: mccurley@almaden.ibm.com
|
| info@www2003.org
Sent by: tbyte@sztaki.hu
02/24/03 10:47 AM
|
To:
Kevin McCurley/Almaden/IBM@IBMUS
cc:
Subject: WWW2003 - Camera-ready
papers |
From tbyte Mon
Feb 24 16: 47 MET 2003
Dear Kevin S.
McCurley,
We kindly inform you that the
information concerning the preparation of
the final camera-ready papers has
been updated on the WWW2003 web site.
The
Permission and Release Form is also available and the deadline of
sending
the form back has been modified. The new deadline is 12 March 2003
and it
should be submitted directly to Sheridan Printing Co., Inc.
(for address
see the form).
Best
regards,
Viktor Richter
WWW2003
Secretariat
|
|
|
Oxford
Brookes University Computer Services
Maintained by WebMail
Administrator
|
|
