W3C Forms teleconference May 27, 2009

* Present

John Boyer, IBM (chair)
Leigh Klotz, Xerox (minutes)
Mark Birbeck, WebBackplane (IRC Only)
Charlie Wiecha, IBM
Nick van den Bleeken, Inventive Designers
Steven Pemberton, CWI/W3C

* Agenda


* Previous minutes

* Next FtF

(Anti)Register: http://www.w3.org/2002/09/wbs/32219/formsabsence2009/

John Boyer: Steven, Erik, and I have filled it out. Check off the calls you'll be absent from. Please do this today.

* XForms 1.1

John Boyer: One of the issues is about upload and serialization, from June. I felt concerned about making changes that said the spec wasn't clear, but the spec looked clear enough to me. There are implementation notes. Please take a look at these by the end of the week to make sure you agree it's OK not to do them:

Action 2009-05-27.1: Erik Bruchez and Leigh Klotz to comment on John Boyer's proposal to rescind actions http://lists.w3.org/Archives/Public/public-forms/2009May/0053.html http://lists.w3.org/Archives/Public/public-forms/2009May/0053.html

John Boyer: The following edits are done: targetid and targetref clarifications; xforms-submit-done and -error on replace=all submissions; the behavior of the copy element in select and select1 tied to setvalue/insert/delete.

John Boyer: The issues on the agenda today are mostly the test suite.

* XForms 1.1 Test suite changes

** W3C license addition

John Boyer: Nick, did you do this?
Nick van: Yes, on the front page only, as we agreed.

** Non-XHTML in test suite


John Boyer: Uli's action here doesn't affect the test suite itself, so we won't hold that up.

** Fixes to six Ch. 5 tests

http://lists.w3.org/Archives/Public/public-forms/2009May/0041.html http://lists.w3.org/Archives/Public/public-forms/2009May/0046.html

John Boyer: The proposed changes are to six legitimate tests; however, they test datatypes, and two of them are base64binary and hexbinary. In those two the upload control is used. It turns out, that for some of the tests, we don't have enough implementations if the uploads were outputs; the focus of the test is base64binary, not upload.
Leigh Klotz: What does output do with base64encoded data in binary?
John Boyer: If it's an image, then it displays as an image. If it's not (text) then it displays the hex.
John Boyer: The tests use events to test the datatypes, and output does that. The events show which types are passing and which are failing. So we would need to change the test.
John Boyer: There are tests of upload in chapter 8, and while Ubiquity doesn't pass them, others do. The Ubiquity processor passes all 20 datatypes, but doesn't implement upload. Others implement upload.
Nick van: It's the same in Chiba.
John Boyer: It's challenging to implement upload in Ubiquity Javascript, but we have some ideas. It's not high priority right now.
John Boyer: The proposal is to change the datatype tests to use output instead of upload. Therefore I move that we change the tests. Does anyone feel that that's a problem? Are you guys ok with that?
Steven Pemberton: I'm ok. You should ask if anyone objects.
John Boyer: Does anyone object? OK.

Action 2009-05-27.2: John Boyer to change XForms 1.1 tests to use output instead of upload: http://lists.w3.org/Archives/Public/public-forms/2009May/0041.html http://lists.w3.org/Archives/Public/public-forms/2009May/0041.html

John Boyer: Great; these six areas are the last that we need to get the Ubiquity implementation report in.

** Fix to Ch. 6 constraint test


John Boyer: We have enough reports, but the test is wrong. We discovered this at the last F2F but haven't updated it. The constraint tests compares two element nodes but doesn't convert them to number. Does anyone object?
Charlie Wiecha: Sounds like a good idea.

Action 2009-05-27.3: John Boyer to add number() to XForms 1.1 test http://lists.w3.org/Archives/Public/public-forms/2009May/0042.html

** Fix three binding exception tests in Ch. 7

John Boyer: The message action listens for the binding exception, and it's listening in the wrong place. They appear to be cut-and-paste errors from the test for compute exception on the model element, but using an output to get a binding exception. The message action is left in the model, but the exception is dispatched to the element. We need to move the actions to pass the test.

Action 2009-05-27.4: John Boyer to fix XForms 1.1 test suite http://lists.w3.org/Archives/Public/public-forms/2009May/0043.html to move message out of model.

John Boyer: In the next one, the message is in head but out of model; it's a little harder for Ubiquity to do this.
Steven Pemberton: Sounds good.

Action 2009-05-27.5: John Boyer to change XForms 1.1 test suite to put binding exception handler inside model instead of outside. http://lists.w3.org/Archives/Public/public-forms/2009May/0049.html

** Deleting xforms-link-error test in Ch. 10


John Boyer: xforms-link-error is gone, but not the test. No objections?

Action 2009-05-27.6: John Boyer to remove xforms-link-error test from XForms 1.1, as xforms-link-error itself has been removed: http://lists.w3.org/Archives/Public/public-forms/2009May/0047.html

** Fix Ch. 11 submission separator


John Boyer: The test is out of date because we changed the default of the separator attribute. I propose that we correct the test to reflect the spec change. Any objections?

Action 2009-05-27.7: John Boyer to change XForms 1.1 test suite to reflect default value of submission/@separator http://lists.w3.org/Archives/Public/public-forms/2009May/0050.html

** Fix Ch. 11 submission final character


John Boyer: The test text says there should be a final semicolon character, but that's not correct according to the spec, or even to the spec example. Any reason it should be a trailing character?
Leigh Klotz: At one point we had some bugs in the spec text that said to have the trailing semicolon, and we fixed it.
John Boyer: So we would get Ubiquity passing now.
Leigh Klotz: We should send a note out to www-forms saying we're changing this test not to require a trailing semicolon or ampersand.
John Boyer: I'd be happy to do that.

Action 2009-05-27.8: John Boyer to fix XForms 1.1 Test Suite to remove trailing semicolon or ampersand as in http://lists.w3.org/Archives/Public/public-forms/2009May/0048.html

Action 2009-05-27.9: John Boyer to send note to www-forms about fixing XForms 1.1 Test Suite to remove trailing semicolon or ampersand as in http://lists.w3.org/Archives/Public/public-forms/2009May/0048.html

John Boyer: That's the last of them. All we need now is the PR spec text.

** Cross-Implementation Report

John Boyer: I will get the Ubiquity implementation report in. Nick, can you run the cross-implementation report after that?
Nick van: Yes.
Nick van: Should we include XSLTForms in the report?
Leigh Klotz: I asked them to send in a report for feature coverage; if we don't need it for feature coverage we don't need to include it.
Steven Pemberton: It's our choice whether to include it in the implementation report or not.
John Boyer: We should put the report into our 2009 space.
Nick van: OK.
John Boyer: Do you have the XSLT for producing the report checked in?
Nick van: test/xforms1.1/implementation-report-generator.

* Next FtF

John Boyer: We need to craft a new charter. Is that a group exercise?
Steven Pemberton: The first step should be to dicuss what we want to achieve. Someone (me) goes off and writes the document that matches those requirements.
John Boyer: What we want to achieve... There's a section now where we discuss version numbers: http://www.w3.org/2007/03/forms-charter.html
Steven Pemberton: We have to be specific with "maintenace" and need to specify "deliverables." Numbers aern't required.
John Boyer: Our current charter adds new deliverables, XForms Transitional, the provisional name for XForms for HTML. We have a FPWD of that.
John Boyer: So we currently have XForms for HTML, XForms 1.2, and XForms 2.0. So we have XForms Refinements, and XForms Next?
Charlie Wiecha: We need to revisit what we mean by 2.0.
Steven Pemberton: This is a good discussion for the face-to-face.
John Boyer: 2.0 is where we put non low-hanging-fruit.
Charlie Wiecha: I'd like to discuss forms vs. web-application. I believe if you deconstruct emerging rich internet applications platforms, you see XForms inside. We have a great asset, an open standards approach to building these applications, but we don't have the charter to do it, nor have we even agreed that it's true.
Steven Pemberton: Are you saying that we should not call it XForms 2.0 but redefine it Declarative Markup for Applications based on the XForms Model, at the F2F?
Charlie Wiecha: Yes, to discuss at the F2F. I'm tired of talking about what forms are and aren't. Let's put it on the agenda.
John Boyer: Let's come up with ideas at the F2F for rechartering and focus.

John Boyer: Let's plan for our Thursday agenda. Is there any more level of detail for the charter?
Steven Pemberton: We should discuss what we want to achieve, and how we want to achieve it, and evaluate solutions. We don't need to go into the details of writing text. That shold be done offline after we agree.
John Boyer: What types of information are needed?
Charlie Wiecha: We also discussed having a conversation with Sam Ruby; it may be that all we do is pass through extension elements, but we should have a discussion.
John Boyer: That's probably not the first step. We need to decide if we're doing XForms for HTML in the new charter.
Charlie Wiecha: And the MVC version of XForms.
Steven Pemberton: Sam Ruby is positive about putting an MVC architecture into HTML, but I haven't had much contact with him.
John Boyer: A Ubiquity-like approach may be the best way to get new ideas into browsers, even for Sam's group to get a majority of browsers to support new language features.
Steven Pemberton: XForms is a good solution, grounded in research and experience, and mustn't be lost for the future of the web.
Leigh Klotz: So you ask Sam to get browsers not to tarnish the DOM, and fix the other problems that get in the way of projects such as Ubiquity.
Charlie Wiecha: We have some notes in the Backplane XG to feed the charter discussion.

* IRC Log


* Meeting Ends