Push Button Paradise

Another chapter online

Chapter 8, covering data submission in all its gory detail. Have a look. -m

A call to action from XForms mastermind Sebastian Schnitzenbaumer:

"Your help is needed."

"No other W3C technology has had so many implementations at Candidate Recommendation."

"The XForms WG is aiming to reach interoperability amoung those
implementations that support the CR spec...Any XForms example that runs in more than one implementation, no matter how simple, is of great value to the working group and the XForms community."

"I'm encouraging everyone to take the time to write XForms...your time and energy used in the right way can now significantly make a difference for the future of web forms"

My sentiments exactly. -m

Software Development 2002

Here's my notes for my presentation at SD2002, Friday the 22nd, at 10:00a. Might not make sense to anyone but me, but having online materials saved me an immense amount of pain at the last conference I presened at.

Easing Development Pain with XForms

Software Development 2002, Boston
Friday, November 22, 10 - 11:30A

Practice Session

What kind of developers is this presentation for?

* Web Interfaces
* Embedded Device with UI
* Remote Administration
* Data Entry / Workflow
* For the Curious and the Thrill-seekers

Demonstration, showing X-Smiles and Novell XPlorer
Main features to show off:
* Document view vs. editor view (bookmarks are XML file, bookmark editor is FO+XForms)
* SMS form, no script needed
* Show off form controls
* Show off calculations
* Show off style
* Show off SVG + XForms
* Calculator

History of Web Interfaces
First, there was no web
Around 1993, forms were added to HTML
nothing new happened for the better part of a decade
Show demo of LiquidOffice HTML
Present day--an explosion of interest

Problems with HTML forms
script, script, script, script :-)
Dynamic assembly required to "prefill"
Primitive, flat, data representation

Scripting Closes Doors
Wouldn't you rather work on the cool parts of your application instead of battling with browsers?

Design of XForms
XML in, XML out
Integrate with XML tools; Web Serivces
Leverage existing XML Technology (XPath, XML Schema, XSLT)
Any time, any place, any device, anybody

XForms Architecture
Building blocks. On one side, the XForms Model, on the other UI and form controls
Close relationship to MVC

Google with XForms:
solves the 'you might be able to just hit enter' specification gap
<input ref=�q�>
<label><html:img src=�images/google.jpg� alt=�Google Search:�/></label>
<send ev:event=�DOMActivate�/>
</input>

Building Blocks
"XForms is not a free-standing document type". Host language required.
head/body separation typical
XHTML, SVG, FO, SMIL, and more

XForms Instance
XForms is a third progeny of XPath, after XSLT and XPointer
Template for data, possibly partial
Supports XML routing and workflow
Datatype support through WXS

Constraints (Model Item Properties)
Static (from Schema) and dynamic
Datatypes
Relevant
Required
ReadOnly
Validation
Calculation
Minimum/maximum occurrences
P3P tagging

Data Submission
XML, or legacy format�
Including Binary data
GET, PUT, or POST
File system, HTTP, or e-mail

Processing Pipeline
Document -> Instance data -> Document

User interface
Expresses intent, great latitude allowed in implementation details
(voice, phone, multimodal, etc.)

Appearance hints
appearance= �full� | �compact� | �minimal�
Much more control coming through CSS

Usability
Help/Hint/Alert integral to the form

Advanced User Interface
Grouping
Dynamic Presentation
Multiple pages
Repeating Tables

Binding
XPath or IDREF -> ID -> XPath

XFormsActions
Defines common set of behaviors, based on XML Events
No JavaScript required
Message
Send
Setvalue
Setfocus
Toggle
Load
Script will still be possible

Unifying client and server code
Same XForms Model that performs client validation can do a server check
Current practice is JavaScript on client, Perl on the server. 2x effort.
Good prevention of the telnet effect (aka. Don't trust the client)
Less CGI code means more security

Section 508
Accessible web forms for everybody
Not keyboard-centric nor mouse-centric
Events for focus, navigation
Script not required
Captions/metadata
Multimodal forms

Roadmap
CR today
Implementation Workshop late Feb 2003
PR April?
Rec May?

Preparing for XForms
Deploy XML data (Web Services)
RPC-style bad; Document-style good; Plain-old XML even better
Think beyond the browser. Heck, think beyond client/server
Avoid dependence on HTML processing
Embrace open standards

Vote For XForms in Mozilla!
http://bugzilla.mozilla.org/show_bug.cgi?id=97806

Implementations/For further info...

-m

RIAA announces Copy-protection System for Compact Discs

Hollywood, CA - The RIAA today announced "the ultimate in protection for the rights of our artists", unveiling the Secure Compact Uber-Disk (SCUD). The technology, which is being tested on the latest Britney Spears album, reportedly is so secure that no player is capable of extracting digital or analog audio from it. This news alone has led to an initial spike in sales, compared to Britney's last album.

As an additional security precaution, the disk has an in-flight range of 150 nautical miles, and carries the destructive yield of 500 pounds of TNT. RIAA insiders have been internally referring to the project as "da bomb".

(now returning to regularly-scheduled boring forms news) -m

XForms Book Building Momentum

Another chapter, filling the gap between 1 and 3, now online. -m

O'Reilly blog: XForms Building Momentum -m

XForms 1.0 is now a W3C Candidate Recommendation. Press coverage:

"The ability to have sophisticated forms so that users can interface with machines more intuitively is perhaps the biggest problem for the non-Microsoft lobby to solve," said James Governor, principal analyst at market research firm RedMonk. (C|Net)

�All of a sudden it creates the one bit of functionality that�s been missing in the XML-based Web, which is interacting with XML documents. It�s inputting and editing data natively in XML,� says Steven Pemberton, co-chairman of the XForms Working Group and a researcher at CWI, the Dutch national research institute. (NetworkWorldFusion)

-m

"To this date in history I have *never* managed to hand-create the RDF for even a fairly simple set of assertions without screwing up at least once." -- Tim Bray

Amen, brother. -m

UIML and XForms from the Cover Pages

UIML is a cool technology, and the basic premise of the comparison--that UIML and XForms have different goals, ambitions, and usages--is sound. The article, however, underestimates XForms somwhat.

The goal of UIML is ambitious: "to provide a meta language powerful enough to subsume the concepts in every language ever designed to describe or implement user interfaces".

The goal of XForms is more mundane, but more practical: to update the nearly decade-old web forms technology.

There's two possible poles of abstraction, namely greatest common denominator (GCD) and least common denominator (LCD). The main problem with LCD is that it defines such a minimal feature set that nobody ends up satisfied with the result. The main problem with GCD is that, in the fully general case, it's never possible to transform from the overshadowing abstract to the concrete. This is the case with UIML.

[In XForms] the UI controls still encapsulate all content of the UI, meaning that label text and UI control content are still integrated into the UI control. This can make internationalization difficult.

This isn't correct. Form control labels can come from separate instance data (through the ref attribute) or even a separate document (through the src attribute).

UIML allows any UI metaphor (e.g., card-based, Web form-based, desktop GUI application, voice dialog, 3D immersive environments, etc.). In contrast, XForms is restricted to the web-forms model of interaction.

As a comparison, it doesn't make sense here to call XForms "restricted". XForms is fully capable of WML-Card-like presentation, phone, voice, multimodal, and yes, the spec even nods to 3D.

UIML should capture author intent.... This view is not really captured in XForms.

XForms is not a meta language, and thus has to draw the line somewhere. "Intent" is the criteria used by XForms, which results in form controls with short, active names like input, select1, and upload. The specific rendering of these (input box, voice prompt, virtual reality, etc.) is entirely up to the specific implementation.

The final, and probably most important, difference is that XForms has run the W3C gauntlet. There's a reason why W3C specs take 2+ years to develop: continuous review by experts in XML, internationalization, style, content, presentation, behavior, bindings, graphics, UI, data formats, metadata, hypertext, schemas, web architecture, web services, and more. While this is far from a guarantee that something good will come out at the end, it has resulted in some nice successes, especially with smaller, more focused Working Groups.

In the end, I see both XForms and UIML being successful in their respective sweet spots, with lots of room for better coordination. -m

XULPlanet lists 101 things that Mozilla can do and IE can't:
(In fairness, IE actually can do several of these)

http://www.xulplanet.com/ndeakin/arts/reasons.html
-m

This is cool. The main blog page is just RSS+CSS. Editing is just RSS+XForms.

Of related interest is the Creative Commons module for RSS. -m

According to this report, EccoPro--the final build of version 4.01--is freely available via FTP. Apparently the poor guys who decided to stop selling it got tired of people asking about it. Maybe this can tide me over until Chandler...

(Thanks to Miguel Marcos for the heads up) -m

Running Code for the Chandler project, with a prototype named "Vista". -m