In particular, I would appreciate it if some folks from the W3C side could check out
their relevant bits and let me know if I represented our work accurately.
The only downside was the attendance. I didn't have my eye on the meter while writing,
but I think there may have been only a dozen or so folks there. Still, a good precedent.
I want to suggest including similar 'town hall' sessions at WWW6; perhaps we can even
begin staging them ourselves when we release a Recommendation or somesuch -- a teach-in
of sorts :-)
Thanks,
Rohit
Anne Papina[Staff] | Welcome to the Internet Publishing Forum
| Today's conference topic is Advancing
| HTML: Style & Substance
| Our guest speaker today is Rohit Khare,
| editor of "Advancing
| HTML: Style and Substance" which is part
| of The World Wide Web Journal
| In this book, W3C offers expert
| introductions to HTML and CSS (Cascading
| Style Sheets) and
| takes a look at
| behind-the-scenes interviews with members
| of the HTML
| editorial review board.
| This book also covers some of the latest
| user interface
| features such as GIF
| animation and JavaScript interactivity as
| well as discussing
| usability
| engineering, accessibility for disabled
| users and multimedia
| technologies.
| If you have a question for Rohit, please
| type a single ? and
| we'll call on you in order.
| Rohit, do you have anything you'd like to
| add?
Rohit Khare | Thanks for the introduction, Anna. I'm
| glad to have the chance to chat about the
| Journal, about some of the technology we've
| covered in this latest issue, and a bit
| about the W3C (the Consortium) and the Journal
| in general.
| This is a tough question to ask on-line,
| but if you had to raise your hands, how many
| of you have heard of the Web Journal,
| or know about the W3C's activities?
Anne Papina[Staff] | Rohit, why don't you start off explaining
| a little bit about W3C
Rohit Khare | OK.
| The Web Consortium was founded by the team
| that invented the Web
| Director Tim Berners-Lee, who developed
| the Web at CERN in Geneva in 1990
| wanted to ensure the open, interoperable
| evolution of Web standards and technology
| so MIT, the European Commission and other
| folks encouraged the concept
| of a vendor-neutral Consortium for
| hammering out specs
| and developing new technology
| We have abot 160 member organizations and
| 35 staff members
| at MIT's Lab for Computer Science, INRIA
| in France, and now, Keio University in Japan
| We are active in a great many areas --
| just scan the jam-packed www.w3.org homepage
| for that! --
Heidrun | ?
Rohit Khare | so we jointly developed the Web Journal
| with O'Reilly as a way to get our ideas out to
| the public and stimulate technical debate.
Anne Papina[Staff] | Ready for a question? :)
Rohit Khare | That, finally, brings us to today's issue,
| which is all about User Interface
| technologies on the Web,
| one of our three domains.
Anne Papina[Staff] | Thanks!
| Go ahead Heidrun
Heidrun | I have a question about the future of
| JavaScript.
Rohit Khare | So do I :-)
| Can you elaborate?
Heidrun | It's about the only thing I know and I
| like it, but I was told to be careful,
| it might not work everywhere
| is it a safe bet, or are there
| alternatives?
Rohit Khare | Yes, maybe, and yes...
| 1) no, it won't work everywhere
| 2) right now, it may or may not be a great
| bet -- depends a LOT on your application
| 3) there are alternatives
Heidrun | what would you recommend as a "safer bet"
Rohit Khare | Of course, those are simplistic answers.
| Here's the scoop as I see it
| ... "a safer bet" depends on what you need
| to do today...
| for example: are you relying on having an
| 'archival' site that will live for years?
Rohit Khare | Or is it a weekly/temporary consumer site
| which is supposed to be flashy?
| The reason I ask is:
| JavaScript, like several of its
| competitors
| (VBScript, Active Server Pages, etc.)
Heidrun | all I really need is a simple user
| interface
| to select documents interactively
| based on a few limited criteria
| JavaScript is working very nicely for
| this.
Rohit Khare | is still in development, and only just
| beginning to be standardized.
| Ah: I see now.
| Are you talking about the common idiom for
| 'menu' lists,
Heidrun | that's right, e.g. by country, name, etc.
Rohit Khare | That's an excellent JavaScript-style
| application
| Limited interactivity that should be done
| at the client side
Heidrun | thanks, I thought so, too.
Rohit Khare | The issue, then, is accessibility to your
| audience
| Of course, some vast percentage of users
| out there
| have the latest and greatest Netscape or
| Microsoft browser,
Heidrun | right again, it is already very
| frustrating
| to see the difference between
| Netscape and Explorer on simple pages
Rohit Khare | but it may be that your audience is
| actually
| in the 'disenfranchised' remainder
| So it's important to have a backup,
| just like a 'non-frames version'
Heidrun | is there a way to check whether a browser
| will do JavaScript?
Rohit Khare | Re: frustration.
| Right now, there's no reliable way to
| check if a browser has feature X
| for X like: sound, Java, big screens, etc.
| Unfortunately, that reduced many advanced
| site designers
| to keying off of the 'User-Agent' by
| keeping track
| of the capabilities of all the hundreds of
| products out there
| (like www.browsercaps.com, for example).
| And it's still not foolproof.
| Architecturally, the Web should handle
| this case
| Just like Content Negotiation can deliver
| a home page in the surfer's preferred
| language or graphics format, 'feature
| negotiation' should
| tell us if a browser has X.
| Right now though, that's only a thought
| experiment in the Web developer community.
| I could say a bit more about the
| standardization process and how ECMA is
| handling JavaScript,
Heidrun | ... leaving us with the smallest common
| denominator, usually not very pretty ...
| please do, uhless there are other
| questions
Rohit Khare | a bit more on the importance of
| accessibilty
| for all browsers , or other questions
| OK, I'll tackle "lowest common
| denominator"
| Standards, by their nature, are inherently
| retrospective
| In this issue of the Journal, we include a
| spec for HTML3.2
| That's the 'latest and greatest' features
| -- as of early 1996!
| It takes a long time to hammer out the
| small details, of course,
| and make political peace along with the
| technology
| Sometimes, it's put as a black-or-white
| question:
| "Do you obey standards?"
| I mean, the implied challenge is: are you
| with-it, or are you hobbling yourself to
| only work with what EVERYONE already
| knows?
| Well, I think it's better to turn it
| around
| and emphasize what standards protect
| developers with
| Standards are the only way to protect our
| investment
| and leverage access to the widest
| audience.
| 'locking yourself in' to a whiz-bang
| proprietary innovation
| is NOT wrong -- it just has to be
| carefully chosen.
Rohit Khare | The much-lampooned <BLINK> tag, certainly
| isn't worth it, of course.
| But what if you're experimenting on the
| cutting edge, with embedded
Rohit Khare | applets (Java, ActiveX, whatever). If you
| dove in early on, you might have
Rohit Khare | used HotJava's APPLET tag.
| But as the HTML process went along, W3C
| sat down with JavaSoft,
| Microsoft, Netscape, and ahost of others
| to merge competing proposals into the
| <OBJECT> tag (published in our previous
| issue, "Building an Industrial Strength
| Web", v1 n4)
| Now, early designers may have to rewite
| their pages and more to make this
| all work for the future. Was that 'risk'
| worth it? Sure -- if it was carefully
| evaluated to begin with
| Finally, sometimes standards ARE
| proactive.
| For example, the Cascading Style Sheets
| spec in this issue.
| The very first Web software ever written --
| over five years ago --
| already had browsing, images, even a
| WYSIWYG editor, and style sheets to
| separate the fonts, linebreaks, etc, from
| the *semantic* HTML markup.
| Over the last few years, the community and
| our staff have developed more powerful
| propsals for Style Sheets. Today,
| designers
| who adopt this new standards can have
| MORE control than they ever did by hacking
| around one vendor's tools:
| precision control of fonts, flow, layout,
| colors, you name it.
| And the architecture is 'cleaner', too:
| you can specify a style for a whole site in
| one fell swoop.
| Most of all, we can now move on to
| 'rendering
| style' for disenfranchised communities.
| We can speak pages with aural stylesheets,
| braille, etc.
| That's how the lowest common denominator
| advances in W3C's view
| Whew.
Heidrun | Talking about styles, can you recommend any
| software for bulk conversions of
| documents to HTML?
Rohit Khare | I don't have any particular packages to
| recommend.
| Of course, it also depends on your source
| document format.
Heidrun | WordPerfect
Rohit Khare | At W3C, we all have to 'eat our own
| dogfood',
| so we only use HTML to begin with:-)
| It's a grand experiment in trying to
| coordinate a fast-moving global team
| using our 'own' Web technology.
| WP->HTML? I don't know anything off hand.
| Of course, the beauty of the Web, and the
| Internet, is that someone, somewhere,
| certainly has.
Heidrun | When you talk about cascading style sheets,
| what exactly is a "style"?
Rohit Khare | I guess your issue is 'bulk', right? That
| may be hard to find.
| Re: CSS.
Heidrun | (_very_ bulky)
Rohit Khare | CSS lets you bind formatting rules to HTML
| elements.
| suppose you have a headline on a page,
| <H1>President Shot</H1>
| That says that phrase is quite important,
| a level one heading. But
| I may want to be more specific. I might
| say this is a political
| headline. So I could use <H1
| <H1 CLASS=Political>
| Now, over in my style sheet, I can say
| "Anything in H1, make it 48 pt.
| Anything specifically H1.Political, make
| Red, and indent it one inch".
| The Cascading part is that I can 'add'
| another style rule, say, from my personal
Rohit Khare | browser preferences, which says "Also make
| anything H1.Political +12 points bigger".
| As for the kinds of things you can control
| today?
| color spacing drop caps columns
| fonts sizes margins
| flow around images
| Coming up soon: z-order (layering in depth)
| and absolute positioning (this image is 1"
| in from the top)
Heidrun | how about a simple TAB at the beginning of
| a paragraph?
Rohit Khare | Again: the moral of the story is reuse and
| protecting the investment.
jim sorensen | ?
Rohit Khare | (yes, first-paragraph indent is OK)
| Today, W3C'
| W3C's online slide shows are actually just
| HTML with a 'slide style'.
Anne Papina[Staff] | Go ahead, Jim
jim sorensen | That's all in html spec? I thought <font>
| was netscape specific
Rohit Khare | Jim: one of the compromises in the HTML3.2
| spec is the inclusion of the FONT tag
| (a matter of some controversy :-)
Rohit Khare | This was accepted because our goal for 3.2
| was to establish a baseline.
| A first step in the process of getting these
| often-fractious members around the table.
BettyM | ?
Rohit Khare | <FONT> may be deprecated in future
| versions as 'for compatibilty only'.
|
Anne Papina[Staff] | Go ahead, Betty
BettyM | Any good book in HTML3.2
Rohit Khare | Betty: I can certainly recommend Dave
| Raggett's "Definitive Guide to HTML3.2"
| from Addison Wesley.
James W. Phillips | ?
Rohit Khare | Dave did a large chunk of the work on that
| spec, working at the W3C.
| But if you're looking for something a little
| less wieldy, I highly recommend "Advancing
| HTML" ;-)
| Seriously, one of the new editorial strong
| points of our Web Journal is that we pair
| up the
BettyM | Learned HTML really fast, is 3.2 easy too?
Rohit Khare | brand-new hardcore technical specifications
| with how-to-guides
BettyM | Thanks
Rohit Khare | So in this isse, we include the 3.2 Spec, a
| behind-the-scenes look at the team which
| created it and
Anthony D. Mather | ?
Rohit Khare | a great overview article by noted HTML author
| Chuck Musicano, "What's New in HTML3.2:
| Formalizing Enhancements to
| HTML2.0"
| Finally: there's really not much new to learn
| in 3.2: it's just evolutionary improvements
|
Rohit Khare | What happens is that smart use of HTML with
| Cascading Style Sheets 1) obviates the
| need for all these
| of picture traffic and 2) it makes the HTML
| more compact by stripping out a lot of the
| format and spacing hacks and
| providing a single central stylesheet.
| Similarly, developing better HTML object
| models (an "html API") for client-side
| scripting (the topic of our next issue),
| can reduce bandwidth by simulating a rich,
| interactive rsponsive page without making
| round-trips
| back to the server to validate every field
| entry or insert your name here.
Heidrun | ?
Rohit Khare |
Anne Papina[Staff] | Go ahead, Heidrun
Heidrun | I have some problems with character sets, in
| particular in French documents,
| how can I make characters like "oe" stick
| and not disappear all the time?
Anthony D. Mather | That's good news. As it stands now, it takes as
| long (or longer) to obtain information on the WEB
| than it did on ARPANET!
Rohit Khare | Character set encodings have been a pain for
| a very long time on the Internet, not just
| the Web.
| You want to be international, but English and
| 8-bit ASCII are so constraining that there
| are too many fragmentary
| interpretations. And that's just for
| positional representations, much less the
| layered glyphs of Arabic, Asian languages,
| or even Latin ligatures and dipthongs, as
| with 'OE'
| In this case, one answer is W3C's
| Internationalization (abbreviated i18n) effort
| to encourage UNICODE characters on the net.
| With 16-bit characters, there is a single,
| unique way to indicate 'oe' without trampling on any
| other languages use of the same id.
| Unfortunately, even in the latest HTML specs,
| the 'ae' dipthong is in the SGML entity set: æ
Heidrun | where would I find a list of &...;
| characters, maybe that would solve my
| problem temporarily
Rohit Khare | but œ is not officially there.
| Appendix D: Character Entities for ISO
| Latin-1, in the HTML 3.2 specification, at
| that Tech Reports (TR) url above.
jim sorensen | ?
Rohit Khare | (this is not HTML's fault: it's ISO-8859-1,
| an effort to cram european languages into
| the 256-position ASCII set. Just
| didn't have room for it, I guess)
|
| Jim:?
jim sorensen | how does unicode compare w/ 16 bit iso
| charsets?
Rohit Khare | Religious wars ahead! ;-)
| Actually, that's a pretty arcane debate, at
| the edge of my ken, so take what I say
| with a grain of salt.
| As I understand it, the original ISO proposal
| was actually a *32* bit space, with
| completely separate
| code pages for several Chinese-derived
| languages.
| by no means 4 billion glyphs (2^32), but
| more than 2^16.
| UNICODE was actually an industry-led reaction
| against the 'wastefulness' of 32-bit
| characters
| that fit all major modern and a great many
| dead languages into 16 bits by sharing
| some common Chinese ideograms. So it's more
| efficient technology vs. more
| go-along-to-get-along nonconfrontational
| international politics.
| W3C is co-sponsoring the next two
| international conferences on UNICODE, so
| perhaps you can find out more about
| this from our i18n pages (in the User
| Interface area of our home page).
|
jim sorensen | ?
Anne Papina[Staff] | Go ahead, Jim
jim sorensen | does the style sheet physically go w/ the
| html code?
Rohit Khare | It can, but need not.
| The css file can be at another URL, using the
| LINK REL=CSS declaration in the HTML text, or
| can be directly encoded in the HEAD.
|
Anne Papina[Staff] | Tell us how we can get a copy of the book!
Rohit Khare | Well, as we wrap things up, I want to
| encourage all of you to amble down to your
| favorite bookstore, real or virtual, and
| pick up a copy of the Web Journal. We're a
| unique publication that can offer insight
| at questions small and large,
| straight from the horse's mouth, so to speak.
| The latest issue, "Advancing HTML" covers
| a slew of advanced UI tips and
| tricks.
Anne Papina[Staff] | The ISBN is 1-56592-264-6 by the way ...
Rohit Khare | Single copies are ~$25, annual subscriptions
| (4 issues/yr) are $75 from O'Reilly and
| Associates.
| Please see http://www.ora.com/info/wj and
| http://www.w3.org/pub/WWW/Journal for more
| information.
| I'd also like to highlight our previous
| issue, which is all about Web Architecture
| and infrastructure, "Building an Industrial-Strength Web"
| and our next two issues, which will be on
| scripting languages and Web Security (Summer '97).
Anne Papina[Staff] | And I'd still like to have you back to
| discuss that book as well :)
jim sorensen | Thank you RK.
Heidrun | thanks a lot, it was very very interesting
Rohit Khare | Thank you Anna, and thanks to all of you for
| participating. I had a wonderful time and
| I'm looking forward to coming back
Anne Papina[Staff] | Thank you Rohit, for all the time & effort
| you put into this
| conference! We certainly had great
| discussion here today!
Rohit Khare | (and just in time too, my laptop batteries
| just gave me the 2-minute warning)
| You're very welcome.
Anne Papina[Staff] | LOL
Rohit Khare | Please, if anyone has any follow up
| questions, do not hesitate to contact me
| directly,
| khare@w3.org.
| Good afternoon... RK
For more information about this and other conference transcripts, please email 74431.2303@compuserve.com.
Copyright 1997 Glenbrook Systems, Inc. All Rights Reserved.