Accessiblitity& standards

3 views
Skip to first unread message

OlivierSeres

unread,
Sep 21, 2005, 3:17:35 AM9/21/05
to TiddlyWiki
I've had an interesting conversation with a W3C representative.
He told me

* TW didn't respect web standards
* TW didn't respect accessibility rules due to AJAX (ie how do
blinds do to read tiddlers ?)
Is it possible to detail these points?
Thanks
-- OlivierSeres

Dale Gillard

unread,
Sep 21, 2005, 7:20:27 AM9/21/05
to TiddlyWiki
OlivierSeres wrote:
> I've had an interesting conversation with a W3C representative.
> He told me
>
> * TW didn't respect web standards

You can check this yourself. Just paste the TiddlyWiki homepage URL
into the W3C validator.

<http://validator.w3.org/check?uri=http%3A%2F%2Ftiddlywiki.com%2F>

While I don't understand why there's a <script> tag outside of the
<html> tag at the start of the document (I imagine there's a good
reason), it seems the W3C validator has problems.

For example, the second error relates to a </div> tag being out of
place. But this Javascript is located in a <script> tag! The validator
appears to be saying that line 673 is part of the <body> tag when it's
clearly not!

It would be good to have TW validate. I'll never be able to use TW on
the intranet in my workplace simply because it doesn't.

Error Line 673 column 52: end tag for element "DIV" which is not open.
this.escapeLineBreaks().htmlEncode() + '</div>';

> * TW didn't respect accessibility rules due to AJAX (ie how do
> blinds do to read tiddlers ?)

I don't think that TW can or should ever respect accessibility rules
relating to Javascript. TW requires Javascript to work. The
accessibility rules limit the use of Javascript unreasonably in my
opinion. The W3C should be asked why they aren't putting pressure on
manafacturers of screen readers, etc to support basic web technologies
like Javascript. Instead the W3C requires everyone to dumb down their
web pages.

Jeremy Ruston

unread,
Sep 21, 2005, 8:18:36 AM9/21/05
to Tiddl...@googlegroups.com
For what it's worth, I'd like to see TW get to the point where it does
validate (the main change will be to the "storeArea" format to use
subelements for metadata instead of custom attributes). Part of the
problem does seem to be (as Dale notes) that current validators are
not very robust.

The accessibility question is more difficult. I'd like TiddlyWiki to
fall-back to a straight static HTML presentation, which seems like the
best approach (and would help things like usage on a primitive PDA),
but it's hard to see how to do that within the other constraints.

Cheers

Jeremy
--
Jeremy Ruston
mailto:jer...@osmosoft.com
http://www.tiddlywiki.com

Daniel Baird

unread,
Sep 21, 2005, 7:19:37 PM9/21/05
to Tiddl...@googlegroups.com

Simon and I were talking yesterday about making tiddlywikis search-engine friendly.. one idea we came up with was to add a static HTML rendered view of some set of tiddlers to the page somewhere (probably at the top..), so search engines could see it sensibly.  Then just eat the entire static display with JS when the page is opened.

We haven't given any thought to how practical the idea is (or at least, *I* haven't given any thought to the practicality of it..) but if it was possible, it would be better for search engines and it would probably be better for screen readers etc too.

Maybe it could be the default set of tiddlers, maybe every tiddler (phew!), maybe a special set called 'staticPageTiddlers' or something, so you could say things like "turn on JavaScript to view this site in it's full glory".

Slightly off-topic, but the preceeding discussion reminded me.

Cheers
Daniel.
--
Daniel...@graduates.jcu.edu.au

carsong

unread,
Sep 22, 2005, 3:08:41 PM9/22/05
to TiddlyWiki
Standardization (sooner better than later)
it will have to pass industry standards before being accepted by the
stream..

Accessibility (benefits everyone).
for deaf people, css has rich screen reading capabilty, for the deaf, a
tiddly could just read itself to the user no? you could even have a
tiddly reading to you in 7.1 channel surround!!

For people without hands, and those who hate their mice... mouseless
navigation -- in the spirit of "wiki" this kinda nav is the target to
reach! think, ... fast ... faster ... terminal velocity!

Someday maybe voice interface ... fastest it will be. Voice +
mouseless navigation would be the fastest possible thing way to
interact with a wiki... until someone writes and API for the brain!

Regarding search engines
aren't the search engys just spiders that crawl the links on a site?
just follow the links. the "javascript:;" maybe can't they just be
changed to reference the tiddler they're gonna hit? (not sure why has
the funny javascript:; link).

Daniel Baird

unread,
Sep 22, 2005, 7:27:34 PM9/22/05
to Tiddl...@googlegroups.com


On 9/23/05, carsong <car...@gmail.com> wrote:

[snip]

Accessibility (benefits everyone).
for deaf people, css has rich screen reading capabilty, for the deaf, a
tiddly could just read itself to the user no?  you could even have a
tiddly reading to you in 7.1 channel surround!!

The main problem with screen reader software and TW (as I understand it) is that screen readers don't run JavaScript.  So all the menus, tags, and other methods of navigation and organisation of your tiddlers are not available to a blind person using a screen reader -- the best the software can do is to read the content of each and every tiddler, in the order they're saved in your HTML file.

Perhaps worse, the screen reader will see the "raw" unformatted text of your tiddler, so the screen reader will be pronouncing all the [[tiddly name brackets]] and //italic slashes// and !!!stuff at the start of headings, and not know that ThisIsATiddlerLink.

I suspect that a large part of the problem is that screen reader software all falls in to the category of stuff that sucks.  Maybe one day Google will bring out "GReader".. but until then, i think if we want screen-reader-accessibility, we might have to do something clever.

And there's plenty of clever folk on the TW lists..


Regarding search engines
aren't the search engys just spiders that crawl the links on a site?
just follow the links.  the "javascript:;" maybe can't they just be
changed to reference the tiddler they're gonna hit? (not sure why has
the funny javascript:; link).


Internal links wouldn't be much of a problem, because the search engine can see all tiddler contents anyway.. but yeah, search engines have much the same problem as screen readers with the formatting.  Try looking at Google's cache of a TW page.. urgh.  Perhaps the biggest problem with TW and search engines is that the title you give your TW isn't seen as a title by the search spider.  So Google etc will get your site name wrong.

Not sure if it's possible at this point to easily solve the search-spider-seeing-my-title problem (apart from editing your TW code directly).

Cheers


D

--
Daniel...@graduates.jcu.edu.au

AlanH

unread,
Sep 22, 2005, 7:32:29 PM9/22/05
to TiddlyWiki
I believe JeremyR fixed the "Google can't see my title" issue by adding
code to the save routine that actually hard-codes your TW title as the
<head><title> title in the html.

Daniel Baird

unread,
Sep 22, 2005, 8:12:44 PM9/22/05
to Tiddl...@googlegroups.com

Really -- that's cool.  This is what I get for just settling down and *using* TW instead of hacking on it.. I'm falling all behind.  Curse the nonTiddly parts of my life!
--
Daniel...@graduates.jcu.edu.au
Reply all
Reply to author
Forward
0 new messages