WebAIM - Web Accessibility In Mind

E-mail List Archives

Thread: Use of <H> tag in PDF

for

Number of posts in this thread: 11 (In chronological order)

From: Chagnon | PubCom
Date: Thu, Aug 29 2013 7:12AM
Subject: Use of <H> tag in PDF
No previous message | Next message →

Trying to get clarification for a client who was instructed by his agency's
accessibility office to use the <H> tag in PDFs, not <H1>, <H2>, etc.

<H> is a predefined accessibility tag in Acrobat, but does anyone actually
use it? If so, are there any drawbacks to using it for all headings or maybe
just the top-level heading or title of a document?

I think the client was misguided and should use <H1>, <H2>, etc., but he's
afraid of getting cited for not obeying his agency's "rules."

Maybe another way to ask my question is: Do you think it is wrong to use <H>
for either the title of a document or all the headings in a document?

Thanks,

-Bevi Chagnon

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - - - - - - - - - -

www.PubCom.com <http://www.pubcom.com/>; - Trainers, Consultants, Designers,
Developers.

Print, Web, Acrobat, XML, eBooks, and U.S. Federal Section 508
Accessibility.

New Sec. 508 Workshop & EPUBs Tour in 2013 - www.Workshop.Pubcom.com

From: Duff Johnson
Date: Thu, Aug 29 2013 7:19AM
Subject: Re: Use of <H> tag in PDF
← Previous message | Next message →

> Trying to get clarification for a client who was instructed by his agency's
> accessibility office to use the <H> tag in PDFs, not <H1>, <H2>, etc.
>
> <H> is a predefined accessibility tag in Acrobat, but does anyone actually
> use it? If so, are there any drawbacks to using it for all headings or maybe
> just the top-level heading or title of a document?

The correct use of the <H> tag in PDF depends on valid nesting of the tag tree. For this reason (among others) I do not believe that <H> is as well-supported by today's AT as are enumerated headings (corrections are welcome).

> I think the client was misguided and should use <H1>, <H2>, etc., but he's
> afraid of getting cited for not obeying his agency's "rules."

No-one's "rules" should require the use of <H> over <H#>; that's absurd. H2, H3, etc. are *certainly* safer than <H> in terms of AT support.

> Maybe another way to ask my question is: Do you think it is wrong to use <H>
> for either the title of a document or all the headings in a document?

Yes.

HTH

Duff.

From: Chagnon | PubCom
Date: Thu, Aug 29 2013 1:09PM
Subject: Re: Use of <H> tag in PDF
← Previous message | Next message →

Thanks, Duff.
My thoughts are similar and I wanted to get another opinion before replying
to my client.
Although our testers could voice <H>, they said it didn't give them enough
of the information they wanted that <H1> does.
-Bevi
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - - - - - - - - - -
www.PubCom.com - Trainers, Consultants, Designers, Developers.
Print, Web, Acrobat, XML, eBooks, and U.S. Federal Section 508
Accessibility.
New Sec. 508 Workshop & EPUBs Tour in 2013 - www.Workshop.Pubcom.com

-----Original Message-----
From: = EMAIL ADDRESS REMOVED =
[mailto: = EMAIL ADDRESS REMOVED = ] On Behalf Of Duff Johnson
Sent: Thursday, August 29, 2013 9:19 AM
To: WebAIM Discussion List
Subject: Re: [WebAIM] Use of <H> tag in PDF

> Trying to get clarification for a client who was instructed by his
> agency's accessibility office to use the <H> tag in PDFs, not <H1>, <H2>,
etc.
>
> <H> is a predefined accessibility tag in Acrobat, but does anyone
> actually use it? If so, are there any drawbacks to using it for all
> headings or maybe just the top-level heading or title of a document?

The correct use of the <H> tag in PDF depends on valid nesting of the tag
tree. For this reason (among others) I do not believe that <H> is as
well-supported by today's AT as are enumerated headings (corrections are
welcome).

> I think the client was misguided and should use <H1>, <H2>, etc., but
> he's afraid of getting cited for not obeying his agency's "rules."

No-one's "rules" should require the use of <H> over <H#>; that's absurd. H2,
H3, etc. are *certainly* safer than <H> in terms of AT support.

> Maybe another way to ask my question is: Do you think it is wrong to
> use <H> for either the title of a document or all the headings in a
document?

Yes.

HTH

Duff.
messages to = EMAIL ADDRESS REMOVED =

From: Michael Moore
Date: Thu, Aug 29 2013 1:34PM
Subject: Re: Use of <H> tag in PDF
← Previous message | Next message →

Just as with HTML screen readers that I have used (JAWS, NVDA, Window Eyes)
allow you to jump to specific levels of headings and in lists of headings
will specify the level giving the user more information about the structure
of the document. Best practice is to use the numbered headings. It is
important to nest heading levels correctly.


On Thu, Aug 29, 2013 at 2:09 PM, Chagnon | PubCom < = EMAIL ADDRESS REMOVED = >wrote:

> Thanks, Duff.
> My thoughts are similar and I wanted to get another opinion before replying
> to my client.
> Although our testers could voice <H>, they said it didn't give them enough
> of the information they wanted that <H1> does.
> -Bevi
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
> - - - - - - - - - - - - - - -
> www.PubCom.com - Trainers, Consultants, Designers, Developers.
> Print, Web, Acrobat, XML, eBooks, and U.S. Federal Section 508
> Accessibility.
> New Sec. 508 Workshop & EPUBs Tour in 2013 - www.Workshop.Pubcom.com
>
> -----Original Message-----
> From: = EMAIL ADDRESS REMOVED =
> [mailto: = EMAIL ADDRESS REMOVED = ] On Behalf Of Duff Johnson
> Sent: Thursday, August 29, 2013 9:19 AM
> To: WebAIM Discussion List
> Subject: Re: [WebAIM] Use of <H> tag in PDF
>
> > Trying to get clarification for a client who was instructed by his
> > agency's accessibility office to use the <H> tag in PDFs, not <H1>, <H2>,
> etc.
> >
> > <H> is a predefined accessibility tag in Acrobat, but does anyone
> > actually use it? If so, are there any drawbacks to using it for all
> > headings or maybe just the top-level heading or title of a document?
>
> The correct use of the <H> tag in PDF depends on valid nesting of the tag
> tree. For this reason (among others) I do not believe that <H> is as
> well-supported by today's AT as are enumerated headings (corrections are
> welcome).
>
> > I think the client was misguided and should use <H1>, <H2>, etc., but
> > he's afraid of getting cited for not obeying his agency's "rules."
>
> No-one's "rules" should require the use of <H> over <H#>; that's absurd.
> H2,
> H3, etc. are *certainly* safer than <H> in terms of AT support.
>
> > Maybe another way to ask my question is: Do you think it is wrong to
> > use <H> for either the title of a document or all the headings in a
> document?
>
> Yes.
>
> HTH
>
> Duff.
> > > messages to = EMAIL ADDRESS REMOVED =
>
> > > >

From: Alastair Campbell
Date: Fri, Aug 30 2013 2:16AM
Subject: Re: Use of <H> tag in PDF
← Previous message | Next message →

On a similar note, what do you do when you run out of headings?

I have been in situations where using an H1 for the document title meant I
ran out of heading numbers further down the document. (It was an in-depth
legal style document.) There I used H for the document title and H1s for
the major sections.

Conceptually, as a PDF is multi-page you could consider it more like a
website than a web page, but I'm still not sure what you'd do with the
document title.

I guess it's not terrible to do that, but are the better solutions?

-Alastair

From: Karlen Communications
Date: Fri, Aug 30 2013 6:00AM
Subject: Re: Use of <H> tag in PDF
← Previous message | Next message →

...but not every document type can be "forced" into the web page model. When
I write books, I don't consider each page of the book a separate page from
the core document requiring repetitive navigation and a link to the next
page. Would this mean that each slide in presentation software would require
the look and feel of a web page, how would this translate for spreadsheet
documents: would each sheet be a "web page or each printed page of the
spreadsheet and how would data be even entered into that design if
repetitive navigation and other web design elements were necessary to force
the model? Documents don't have domains and there are different document
types that need to be used for different types of content.

Not everything is a web page and we can't start thinking that everything is
and try to get it to conform to web design. I agree with Birkir's comments
on desktop published documents but would ask why, after how many years of
knowing that access to content is important for those who need or want to
access content differently/using what we currently call adaptive technology,
that students aren't taught how to make those award winning designs
accessible and why the software to let them do that lags so far behind the
need.

Cheers, Karen



-----Original Message-----
From: = EMAIL ADDRESS REMOVED =
[mailto: = EMAIL ADDRESS REMOVED = ] On Behalf Of Alastair Campbell
Sent: Friday, August 30, 2013 4:17 AM
To: WebAIM Discussion List
Subject: Re: [WebAIM] Use of <H> tag in PDF

On a similar note, what do you do when you run out of headings?

I have been in situations where using an H1 for the document title meant I
ran out of heading numbers further down the document. (It was an in-depth
legal style document.) There I used H for the document title and H1s for the
major sections.

Conceptually, as a PDF is multi-page you could consider it more like a
website than a web page, but I'm still not sure what you'd do with the
document title.

I guess it's not terrible to do that, but are the better solutions?

-Alastair
messages to = EMAIL ADDRESS REMOVED =

From: Duff Johnson
Date: Fri, Aug 30 2013 7:04AM
Subject: Re: Use of <H> tag in PDF
← Previous message | Next message →

> On a similar note, what do you do when you run out of headings?

<giggle> Ah yes - one of the premiere showcase examples of HTML being broken by definition.

What genius decided there exist only 6 levels of headings in all of creation? (Sorry Tim!)

Sadly, the same thinking (to a lesser extent) infected PDF back in the day as well. While PDF/UA provides for unlimited heading levels the full solution in PDF will have to wait for PDF 2.0 (ISO 32000-2) which will add H7, H8, H9, etc. (without limit) to PDF itself. PDF 2.0 will also include a <Title> tag, thus providing a definitive way out of the undesirable practice of using H1 for the Title.

> I have been in situations where using an H1 for the document title meant I
> ran out of heading numbers further down the document. (It was an in-depth
> legal style document.) There I used H for the document title and H1s for
> the major sections.

Titles are not structural elements in a document. The document's title should be placed in the document's metadata "Title" field; it's fully accessible to AT in this context without reference to headings (or heading levels) at all.

H is not 'above' H1; there's no reason to think that AT would (or could) do the right thing in this case. What does H before H1 mean in structural terms? I have no idea. This "solution" is a hack that (if it works today on some AT) likely relies on some behavior that's going to change once the AT / PDF reader in question supports standards for accessible electronic documents (PDF/UA) and ISO 32000-2 (PDF 2.0).

> Conceptually, as a PDF is multi-page you could consider it more like a
> website than a web page

Nah. It's a document - not a "page" and not a "website".

There were documents long before there were websites; there will be documents long after websites have been replaced with HyperText Clouds or whatever other name we'll use in 20 years time.

Trying to squeeze documents into the conceptual framework of the web is one of the great goofs of the HTML-centric accessibility world.

> , but I'm still not sure what you'd do with the
> document title.

Place the title in the document's metadata; any PDF-aware AT readily finds it there. Tag title text appearing on the page with a <P> tag (unless it is also serving a structural role in addition to being a title).

Duff.

From: Alastair Campbell
Date: Fri, Aug 30 2013 7:38AM
Subject: Re: Use of <H> tag in PDF
← Previous message | Next message →

Karen wrote:
> Not everything is a web page and we can't start thinking that everything
is
> and try to get it to conform to web design.

Apologies, I wasn't trying to fit a document into a webpage model, just
trying to work out what constitutes an H1.

In the context of a long document it does not seem wrong to use multiple
H1s, but it does in web pages.


Duff Johnson wrote:

> What genius decided there exist only 6 levels of headings in all of
> creation? (Sorry Tim!)
>

To be fair, it's never been an issue for a website I've worked on, only
PDFs, and then only big, lawery type ones.



> PDF 2.0 will also include a <Title> tag, thus providing a definitive way
> out of the undesirable practice of using H1 for the Title.
>

Ah, good to know.


There were documents long before there were websites; there will be
> documents long after websites have been replaced with HyperText Clouds or
> whatever other name we'll use in 20 years time.
>

I agree, there will always be longform content in one file/location. I
personally hope it doesn't require being a print/page oriented format
though. (Although eBook formats aren't a replacement yet either.)

I'd be interested to know if PDF 2.0 requires pagination?


Place the title in the document's metadata; any PDF-aware AT readily finds
> it there. Tag title text appearing on the page with a <P> tag (unless it is
> also serving a structural role in addition to being a title).
>

Thanks, I'll remember that for future. It is primarily for cover-sheet
style pages where you have a title in very big text, it feels like a
title/heading tag would be appropriate, but in the context of the PDF tags
it isn't.

Cheers,

-Alastair

From: Duff Johnson
Date: Fri, Aug 30 2013 1:03PM
Subject: Re: Use of <H> tag in PDF
← Previous message | Next message →

> Apologies, I wasn't trying to fit a document into a webpage model, just
> trying to work out what constitutes an H1.
>
> In the context of a long document it does not seem wrong to use multiple
> H1s, but it does in web pages.

So, I'd like ask the HTML gurus a question hereĀ…

Since web pages already have Title tags why is H1 so commonly press-ganged into the role of an ersatz Title tag? Does this not make the page's true title ambiguous? What's the point / value of this practice other than to screw things up and effectively a heading level from the 6 available?

Oh, and why is multiple H1s a problem? I get the idea of custom, but what's the actual *reason*?

> To be fair, it's never been an issue for a website I've worked on, only
> PDFs, and then only big, lawery type ones.

Fair enough, but why preclude deeply structured content in HTML? What was / is the point? Why wasn't it fixed for HTML5? Etc.

> I agree, there will always be longform content in one file/location. I
> personally hope it doesn't require being a print/page oriented format
> though. (Although eBook formats aren't a replacement yet either.)
> I'd be interested to know if PDF 2.0 requires pagination?

At its core, PDF technology will remain fixed-layout for all sorts of good reasons. With tagged PDF, however, that content becomes redeployable for different readers and by AT.

Since it will always be fixed-layout at its core PDF will also - by definition - be paginated. Note that this is *not* the same thing as saying that it's print-oriented. A page that's 1 kilometer by 1 kilometer probably isn't for printing.

In PDF a "page" can be very very large or very small and a very extreme rectangular shape (if you want), but it will be a page. It need not have anything to do with "print" per se.

Duff.

From: Alastair Campbell
Date: Sat, Aug 31 2013 2:09PM
Subject: Re: Use of <H> tag in PDF
← Previous message | Next message →

Duff Johnson wrote:

> Since web pages already have Title tags why is H1 so commonly press-ganged
> into the role of an ersatz Title tag? Does this not make the page's true
> title ambiguous? What's the point / value of this practice other than to
> screw things up and effectively a heading level from the 6 available?
>

There is some duplication, but they do serve different purposes.

The title is for more than just the page you've opened, it may be your
first interaction with the site, so it often contains more information than
the main heading of the page. (Imagine landing on page 26 of a PDF, where
the first heading of that page and the title would be different.) Its also
used by search engines for the link to the page.

Typically we recommend a title includes:
Page name | Site name | Section name

The 'main' heading would only have the page name.

That's really where "only have one H1" comes from, there is generally a
main heading for (a web) page, visually. It's generally at the top of the
content area, so becomes a good anchor point for people (Especially screen
reader users to skip to. In Jaws, press 1.).


Fair enough, but why preclude deeply structured content in HTML? What was /
> is the point? Why wasn't it fixed for HTML5? Etc.
>

I don't know why it was limited to 6 levels, I guess the new HTML5
(levelless) algorithm would not have that restriction, but it isn't
backwards compatible so I don't see it working for a while.

Given the ability to break a web page into multiple pages, getting down to
H6 is pretty rare unless you are making something very large (even the WCAG
2 doc only uses H1-3).

-Alastair

From: Jennifer Sutton
Date: Sat, Aug 31 2013 3:21PM
Subject: Re: Use of <H> tag in PDF
← Previous message | No next message

Comments below, beginning with JS.
At 01:09 PM 8/31/2013, Alastair wrote, in part (when discussing the
use of title and h1:


>There is some duplication, but they do serve different purposes.
>
>The title is for more than just the page you've opened, it may be your
>first interaction with the site, so it often contains more information than
>the main heading of the page. (Imagine landing on page 26 of a PDF, where
>the first heading of that page and the title would be different.) Its also
>used by search engines for the link to the page.
>
>JS: It's also used when creating a bookmark/favorite, so often, the
>title and page heading should differ.


And increasingly, reading app.s like Instapaper/Pocket use this kind
of information, so it's very helpful to have both title and h1 be
specific in other (somewhat newer) contexts.

>Alastair: Typically we recommend a title includes:
>Page name | Site name | Section name
>
>The 'main' heading would only have the page name.
>
>JS: This seems most sensible to me. Thanks.


Jennifer