Details and Opinions About PDF Accessibility – A Record Aside

PDF information on the net are
generally annoying and fairly often pointless. However after they aren’t both of these issues, we have to make them accessible for a similar causes we make different internet content material accessible.

Article Continues Under

Opposite to common opinion – and in addition opposite to quasi-judicial claims in some locations – PDF paperwork could be no much less accessible than HTML. Whereas this can be a surprising revelation, it’s nonetheless true. This text will clarify how PDF does and doesn’t help accessibility.


  1. Most PDFs on the net ought to be HTML.

  3. Some paperwork actually ought to be PDFs.

  5. You possibly can add XML-like tags to offer construction to a PDF.

    1. Tags weren’t accessible till a latest improve to the PDF file format.
  6. Most display screen readers in widespread use can learn PDFs.

    1. Display screen readers needed to be upgraded to grasp tags.

    3. Display screen readers have been repeatedly up to date all through their historical past, and even as we speak some display screen readers can not deal with elements of the HTML spec.
  7. Even an untagged PDF could be accessible when you’re utilizing the best know-how.
  8. Posting a PDF on-line with no HTML various doesn’t robotically represent discrimination.

Let me dedicate this dialogue to 2 nations that battle beneath the yoke of lies and misunderstandings regarding PDF accessibility.

  1. First, to the individuals of Australia, whose federal human-rights physique, the Human Rights and Equal Alternative Fee (HREOC), takes the official place that posting a doc on-line solely as a PDF is inaccessible, therefore a violation of the Incapacity Discrimination Act.

    I realized this the laborious method from Bruce Maguire. He vanquished Juan Antonio Samaranch in a listening to with that selfsame HREOC over the inaccessibility of the Sydney Olympics web site, and he now works for the HREOC. Maguire monologuized endlessly concerning the matter once I met him in 2004. Once I bought a phrase in edgewise, Maguire was capable of agree, as most of us will, that HTML is the popular format, however he additionally instructed Microsoft Phrase as a substitute for PDF. This may occasionally inform you all you could learn about HREOC’s understanding of accessibility and interoperability.

    I instructed Maguire then and I’m telling everybody now: Don’t consider the HREOC if they arrive after you claiming your PDFs are inaccessible, therefore unlawful. In the event you created your PDFs incorrectly, HREOC could also be proper, however solely generally, and there are a lot of circumstances the place your PDFs could also be simply fantastic.

    In the event you get hauled in entrance of HREOC for “unlawful, inaccessible” PDFs, think about this text a case for the case for the protection, to paraphrase Christopher Hitchens. (See dialogue under.)

  2. Second, to the individuals of Canada, the place the federal authorities’s tips for its personal web sites – recognized euphemistically as Widespread Look and Really feel, and the supply of nice ache – state falsely that PDF is “in a roundabout way accessible to individuals with (primarily) visible impairments” and that solely “minimal model 2.1” ought to be used. Oddly, there is no such thing as a such factor as a PDF model 2.1.

    The Alliance for the Equality of Blind Canadians, a foyer group that undoubtedly isn’t the Canadian Nationwide Institute “for” the Blind, handed a decision at its 2005 annual assembly stating that “Transportable Doc Format (PDF) continues to supply limitations to blind, deaf-blind and partially sighted individuals and their enabling applied sciences; herefore, be it resolved that the AEBC advocate that PDF not be used as an ordinary for offering paperwork on all web sites.” Nicely, who was suggesting that within the first place?

A view of the panorama#section4

Let’s start originally and talk about all the panorama of PDF on the net earlier than we study what makes a PDF accessible or inaccessible.

The very first thing to do with a PDF#section5

…is Google the URL. Severely. More often than not, Google does a half-decent job of creating a PDF readable in HTML. Poor character encoding, ill-constructed multicolumn PDFs, and doc security measures can forestall Google from indexing a PDF in any respect, or doing so readably. Nonetheless, that’s what I do first.

PDF is overused#section6

There are too many PDFs on the net. Most each PDF ought to be one thing aside from PDF. Any easy text-and-graphics doc that’s typeset in a single column ought to be offered as an strange HTML+CSS+JavaScript internet web page. Actually, I can’t consider any exceptions for easy paperwork.

There aren’t many classes of on-line doc that actually ought to be PDFs and nothing else. And the checklist has decreased by one within the final 12 months, since presentation slides can now be adequately dealt with by Eric Meyer’s S5 technique. (Therefore I’ve no excuse anymore for publishing my very own presentation slides in PDF, so I’m going to cease.)

But when your doc is among the following, PDF could also be fantastic:


  1. Footnoted, endnoted, or sidenoted, since there is not any method to mark up any of these buildings in HTML. (You should utilize a hack like sub or sup for the footnote reference, however there are no footnote, endnote, sidenote, and even notice parts. That hack could also be enough for easy footnoted paperwork, however strive rendering David Foster Wallace’s footnotes-within-footnotes in HTML 4.)

  3. An interactive kind, since PDF interactivity can do greater than HTML can. (Use with warning and provided that HTML actually can not do what you need.) For examples, test Jeremy Tankard’s order varieties, particularly for TypeBookOne (PDF).

  5. A multimedia presentation, since later variations of PDF can actually embed multimedia somewhat than merely confer with or name multimedia, as HTML does. (Similar warning as above.) PDF multimedia can embody captions and/or audio descriptions.

  7. Mixed accessible and inaccessible variations. A typical case is a scan of a historic doc that additionally consists of dwell textual content. (You actually need that dwell textual content. The Smoking Gun’s scanned court docket paperwork wouldn’t go muster right here.) One other instance – one that’s authorized in Canada beneath a copyright exemption – is a sign-language translation inside or alongside a written textual content or audio recording.

  9. Customized-crafted solely for printing. I actually imply that, and never a doc so badly designed that individuals haven’t any alternative however to print it out as a result of studying onscreen is so tedious. Your service-bureau information, if they’re on the net in any respect, can keep PDFs.

  11. Designed for annotation and round-trip journey: In the event you’re posting one thing to elicit feedback, that are then despatched again to you, PDF has helpful buildings that HTML doesn’t.

  13. A sort specimen, that are all however unimaginable to create in HTML, until the specimen concerned is a “typeface” like Arial.

  15. A pattern of a format that can’t be rendered in a browser (e.g, Illustrator or Photoshop paperwork) or can solely be rendered unsatisfactorily (CAD drawings the place GIF and JPEG don’t have sufficient decision). (In concept you might use SVG for CAD, however SVG stays largely theoretical, doesn’t it?) This case additionally consists of PDF information meant as samples of PDF information.

  17. A report of a doc’s state at a selected second. On this context, PDF is beneficial as a preservation format even for HTML internet pages.

  19. A doc in a language whose script has no passable help in internet browsers. This instance have to be used with warning: In 2005, there aren’t many “minority” languages that can’t be rendered in a browser. Maybe this case have to be restricted to scripts that haven’t been encompassed by Unicode (of which there are a number of). This may also be a subset of the type-sample case in case your PDF is supposed as an illustration or documentation of the writing system utilized by a language.

  21. Mathematical, since even MathML can not render sure notations.

  23. Paperwork with a legally restricted format, like U.S. tax varieties.

  25. Paperwork with digital rights administration, which everyone hates and which has doubtless accessibility limitations. (The usage of 128-bit encryption with PDF is suitable with display screen readers.)

  27. Multicolumnar, significantly if figures and illustrations are included, since multicolumn internet layouts are a mere hack and are unreliable as a way of reproducing print layouts. (Your multicolumn doc ought to be HTML whether it is introduced that method merely to save lots of paper and it could actually work as a single column. It may be tough to differentiate that case from a doc that’s structurally multicolumnar, and this class is considerably iffy.)

So let me say the identical factor yet another time: In case your doc isn’t a type of or isn’t actually distinctive ultimately, it’s your duty to do what the remainder of us did and learn to use HTML, CSS, and JavaScript accurately. Use these applied sciences until you may have an hermetic purpose to not.

PDF isn’t Acrobat, and even Adobe#section7

Let’s get one thing else out of the way in which: Acrobat isn’t PDF and PDF isn’t Acrobat. Many applications can show PDFs aside from Acrobat. GSview is a well-liked alternative and works on Home windows, Mac, OS/2, and Linux. Different choices:

  • On Home windows: Jaws PDF Editor (not the display screen reader)
  • On Mac OS X: Preview, GraphicConverter, Safari, and OmniGraffle
  • On Linux or Unix: OpenOffice
  • On PocketPC and Home windows CE: Primer PDF Viewer
  • On Symbian: PDF+
  • On Amiga (!): Apdf

And plenty of applications can create PDFs aside from Acrobat. Almost each utility on Mac OS X can save a PDF, for instance. Utilities to export to PDF with out Acrobat on platforms like Home windows and Linux are too quite a few to say, however websites like VersionTracker checklist them. Furthermore, you don’t even want utility software program to create a PDF; some elite builders instantly write their very own native PDF information.


Your Acrobat model quantity and the PDF file format model quantity are two various things. It could come as a shock that PDF really has a model quantity, but it surely does. There is no such thing as a single “PDF” format any greater than there’s a single HTML format.


  • The most recent model of the Adobe PDF file format is Model 1.6 (launched November 2004).

  • Every new model has added a number of options, a lot of them structural. PDF tags, that are somewhat vital for accessibility, had been added in PDF 1.4.

  • PDF variations for archiving (PDF/A) and “alternate of print-ready pages”  (PDF-X1a) are already ratified or within the technique of ratification by requirements our bodies.

  • There’s a working group to outline an “accessible” PDF format (I’m on it) and one other for engineering.

The simplest method to hold issues straight is to take the present Acrobat launch quantity, subtract 1, and put the consequence after the decimal level. Acrobat 7 can learn PDF 1.6 paperwork, for instance.

Proprietary vs. open#section9

You possibly can’t make the explicit assertion that PDF is “proprietary” and HTML is “open.” The World Broad Net Consortium copyrights its specs, for instance, although with cheap utilization phrases. Adobe publishes its variations of the PDF format, and has carried out so since Model 1.3.

I’ve by no means understood the objection that Adobe may change the PDF format in a single day and render your paperwork ineffective. That objection applies to the mysterious Microsoft Phrase file format, however not right here. (Phrase’s XML schemas have been revealed, however I’m not speaking about these. Microsoft’s PR apparatchiks within the U.S. and Canada promised to get again to me concerning the precise state of disclosure of the Phrase file format, however by no means did.) PDF specs are revealed, and any of your paperwork that adjust to the revealed spec will stay unchanged when the spec is up to date. Simply as a doc validated in opposition to HTML 3.2 remained unchanged after XHTML 1.1 got here out, your PDF 1.4 paperwork (for instance) will proceed to work into the indefinite future.

All the dialogue of proprietary vs. open is bogus. The related distinction is between revealed and secret. PDF and HTML are each revealed codecs. Finish of story.

A number of codecs should be accessible#section10

The aim of the accessibility advocate is to enhance accessibility for individuals with disabilities, interval. We’re not curious about making solely HTML internet pages accessible. The whole lot of internet content material is our purview, and that features codecs like PDF and certainly Flash. (Similar goes for multimedia.)

To attract a historic analogy, after we in Canada and the U.S. managed to get one or two open-captioned TV reveals on the air within the Nineteen Seventies, we didn’t cease there. We invented a closed-captioning system that might be utilized to all applications (and the Europeans reused their present teletext system for a similar function). Then we made positive that residence movies and laserdiscs (keep in mind these?) had been captioned. Then we began releasing a only a few prints of open-captioned films.

Then we found out a method so as to add audio descriptions for the blind to tv applications. Then we hacked a way to provide described residence movies. Then we developed closed-captioning and -description techniques for first-run films (and new techniques for open captioning there). Then we used closed captions, subpictures, and audio tracks to make DVDs accessible. Then we developed strategies to caption and describe on-line video. We saved up with know-how and made every new format accessible.

Even when you by no means create PDFs your self, I’m positive you’ll admit that it’s mandatory for this widespread format to be accessible for a similar causes we made the widespread format of HTML accessible.

We’re not simply speaking about blind individuals#section11

And remember that accessibility isn’t about making issues work fantastic for blind individuals and nobody else. Everyone falls into that entice at one time or one other – the Net Accessibility Initiative included. In PDF accessibility, notable further teams embody deaf and hard-of-hearing customers and folks with studying disabilities. Motor or dexterity impairment turns into a problem in scrolling a PDF.

Preserve this in thoughts the subsequent time somebody complains {that a} sure PDF is inaccessible as a result of she or he couldn’t get it to work with a sure model of Jaws. We’re not working only for you.

Content material vs. consumer agent#section12

To complete this preliminary dialogue, we have to perceive the interplay between content material and “consumer brokers,” the latter being a time period for browsers, media gamers, and different gadgets that current content material to the consumer.  As internet authors, we’re so involved with HTML and CSS that we both overlook the function of the consumer agent fully – or we overlook it till it bites us within the arse, as with CSS bugs in browsers. We have a tendency to not discover the truth that internet content material is (in virtually all circumstances) rendered by a browser; the consumer agent turns into invisible.

However as a result of we’ve to change to a different program more often than not to learn a PDF, we’re instantly reminded {that a} consumer agent is definitely in play. This, too, is a supply of confusion between Acrobat and PDF. The interplay of consumer agent and PDF content material is as vital as it’s with HTML, as we’ll see; it’s simply that we’re extra acutely aware of that interplay.

The grievance that it’s a must to use a “particular program” to learn a PDF doc is bogus. You’re already utilizing a particular program to learn an HTML doc. It’s simply that you just use that program a lot it now not appears particular.

As with HTML, what makes PDFs “strong,” reformattable, and in any other case accessible to many individuals with disabilities is construction. An unstructured information format like a JPEG image is difficult to make accessible, not less than to a blind individual, however wrap it within the HTML construction of an img factor and instantly accessibility turns into an actual choice.

A PDF is a database of various information sorts. You possibly can embody a variety of textual content, graphics, and multimedia codecs inside a PDF, a incontrovertible fact that led to a standard misunderstanding that PDFs are glorified photos. (They actually could be, however they aren’t essentially.) There actually was no such factor as a construction to PDF till tags had been launched in PDF 1.4. It’s OK to name them tags and never parts.

PDF tags are XML-like and shall be instantly comprehensible to anybody with HTML information.  Many tags are functionally equal to analogues in HTML, akin to P, headings (together with a generic, unnumbered Heading factor), and Determine (picture). However a few of these tags have extra options than their analogues in HTML. For photographs, you’ve bought three ranges of alternative textual content – “precise textual content,” helpful for textual content rendered as a picture, a drop capital, or an illuminated manuscript; “alternate textual content,” precisely as in HTML; and “title,” additionally as in HTML. You possibly can and nonetheless ought to declare a language to your PDF doc, simply as with HTML.

PDF tags are extensible and you may create your individual. Nevertheless, there’s a predefined set.

A key distinction right here is that you just can not simply hearth up a textual content editor so as to add tags to your doc, as you possibly can with (X)HTML. Presently, you could use utility software program, very a lot together with Acrobat, so as to add the tags; as a result of PDF is a binary format, that’s unlikely to vary.

As with semantic HTML, a tagged PDF could be reused and reformatted, for the reason that utility software program is aware of what you meant by the information within the doc. It is aware of that this textual content is a headline and this different textual content is a paragraph, so the software program can, for instance, reflow your textual content from a two-column to a single-column doc. Reflow turns your multicolumn PDF right into a zoom format.

Normally, we are able to say {that a} PDF might be accessible whether it is tagged. As with HTML, you possibly can tag issues improperly or unsemantically, although there is no such thing as a idea of legitimate tagging with PDF. The mere presence of tags doesn’t assure accessibility, since you could be utilizing them incorrect, however the absence of tags ensures that the PDF itself isn’t accessible. Notice the emphasis on itself; this isn’t the tip of the story.

The consumer agent’s job#section14

Right here is the place the consumer agent is available in. Simply as internet browsers have needed to be engineered, at huge effort and expense, to deal with tag-soup HTML, Acrobat particularly has needed to be engineered to deal with real-world PDFs. It’s a much bigger drawback, since few PDFs are tagged and the free-form database construction of PDF lacks even the quasi-structure of tag-soup HTML.

On this case, the consumer agent overcomes the inaccessibility of the content material, and that’s the way it ought to be even with HTML: All the chain from writer to reader must be accessible, and any hyperlink within the chain can take up the slack. In accordance with the traditional recommendation to be strict in what you produce and lenient in what you settle for, if the PDF writer creates inaccessible content material, the reader software program ought to attempt to repair it.

  • Acrobat variations since 4.05 (with a Home windows-only plug-in) have been not less than adequately competent a number of the time in making “inaccessible” PDFs functionally accessible.
  • Acrobat 5 and later can infer a studying order and reflow textual content.
  • Acrobat 6 and later can learn textual content out loud on Home windows and Macintosh, functioning as a de facto display screen reader.

Notice that this dialogue largely pertains to blind or learning-disabled readers. A deaf individual would possibly simply learn the doc with no bother. Mobility or dexterity impairment can be concerned right here; later Acrobat variations can autoscroll a doc with out having to tediously click on or actuate a scrollbar utilizing your sluggish adaptive know-how.

Thus it’s unimaginable to state that even an untagged PDF is inaccessible. Acrobat (or another program) might be able to use synthetic intelligence to hack its method by a doc to make it adequately accessible. If somebody complains that your PDF isn’t accessible, you could ask them what program they’re utilizing to learn it. Provided that Adobe Reader (né Acrobat Reader) is free for Home windows, Mac OS, and Linux and has all of the accessibility options listed above, and is, furthermore, suitable with many display screen readers, it’s a little bit of a stretch to say that an untagged PDF couldn’t presumably be learn accessibly. Some PDFs shall be inaccessible even with a very good studying program, however many will work adequately properly.

One other recurring grievance in PDF accessibility –  additionally bogus – is that display screen readers both can not deal with PDFs or require pricey upgrades to deal with them.

All main display screen readers in use on Home windows can learn PDFs, together with Jaws, Window-Eyes, IBM Dwelling Web page Reader, and Hal.

Keep in mind Bruce Maguire? His presentation on the Net Necessities 2004 convention in Sydney – in whose lunchroom he and I talked – acknowledged the next:

The PDF format has change into broadly used for making paperwork accessible on Net pages. Regardless of appreciable work carried out by Adobe, PDF stays a comparatively inaccessible format to people who find themselves blind or vision-impaired. Software program exists to supply some entry to the textual content of some PDF paperwork, however for a PDF doc to be accessible to this software program, it have to be ready in accordance with the rules that Adobe have developed. Even when these tips are adopted (and there are 32 pages of them), the ensuing doc will solely be accessible to these individuals who have the required software program and the abilities to make use of it. Many blind or vision-impaired individuals do not need the monetary freedom to spend the $1,000+ sometimes required to improve their screen-reader software program to reap the benefits of the newest accessibility options. Requiring a consumer to improve to this extent in an effort to learn an ordinary doc is like designing Net content material presentation in such a method that most individuals should purchase a brand new laptop in an effort to learn it. Clearly, this isn’t an inexpensive strategy to the discharge of a authorities’s social duty to supply related data to its residents. In any case, a number of the PDAs utilized by blind individuals haven’t any services for accessing PDF information.

Let’s unpack these objections.


  • Making ready PDFs “in accordance with the rules that Adobe has developed” is on no account totally different from making ready HTML pages in accordance with the rules the W3C has developed. (Wish to print these out? It’s a 34-page PDF.)

  • It’s not like PDFs are the one merchandise in your laptop for which you require software program and expertise. You require each of these to surf the net and use HTML pages.

  • It’s false to say that blind persons are “sometimes required” to pay “$1,000+” to improve their display screen readers. Some gadgets that may learn PDFs aloud are free, like Adobe Reader. No one is requiring an costly software program improve.
  • Let’s have a look at the improve costs for Home windows display screen readers. Assumptions: All costs in U.S. {dollars}; you already personal a replica of a display screen reader that can’t learn PDFs (more and more unlikely).

    1. A Jaws for Home windows Software program Upkeep Settlement offers you the 2 subsequent releases for $180 or $260. Therefore an improve to a PDF-capable model could be “free” beneath this plan.

    3. An identical scheme supplies for 3 upgrades of Window-Eyes for $299.

    5. Upgrading Dwelling Web page Reader from any Model 2.5 or 3 launch to Model 3.04 is free. In any other case you purchase the entire package deal once more for a reduced value of $79.

    7. Upgrading from V5 to V6 of Hal prices $160 or $220.
  • “PDAs utilized by blind individuals” should be upgraded in the event that they don’t perceive PDF. Basically, this objection boils right down to “if it doesn’t work with what I’ve already bought, it doesn’t work, interval.” I suppose time doesn’t march on for these individuals. In that case, I hope you’re having fun with HTML 2.0 and your Geocities homepage.

The idea appears to be that blind individuals solely ever or solely can or solely should use Home windows, and, as everyone knows, Home windows display screen readers are overpriced. Nicely, sure, they’re, however you the blind individual have many choices now for laptop accessibility.

After all you’ll have to replace your Home windows display screen reader and after all that may price you cash. You possibly can’t complain that PDF is inaccessible (because it was for a few years) after which act as if the issue hasn’t been addressed. Adobe rewrote the PDF spec to incorporate tagging for accessibility, and, simply as with every improved know-how, your display screen reader needed to be upgraded to deal with options that by no means existed earlier than. You requested for one thing new and also you want one thing new to make it work.

More often than not once I run throughout this grievance, it strikes me as a peevish try and cling to the discredited concept that PDF isn’t accessible and Adobe (particularly) doesn’t care about the issue. Actually, we want these complainers to develop up and face information.  You possibly can’t ask for a format to be upgraded to incorporate accessibility after which complain that your individual software program must be upgraded.

In the event you don’t wish to fork over the cash for a Home windows display screen reader, you should use Mac or Linux. VoiceOver on Mac OS X 10.4 Tiger can learn PDF 1.5 information and earlier, although not all the time very properly. The Solar accessibility package deal for Linux (a part of Solaris 10), which is freed from cost, consists of built-in display screen studying. There’s now a model of Adobe Reader 7 for Linux, although it doesn’t have speech output.

In the event you’re actually involved about price, set up the free Linux software program or set up Tiger on a used Mac. (Really, a brand new Mac Mini with out monitor prices lower than a brand new license for Jaws.) In the event you suppose that Home windows-screen-reader makers are overcharging, complain to the makers or vote together with your ft. The precise subject – PDF accessibility – is being dealt with.

Let’s be constant about screen-reader flaws#section16

Additionally, when you’re going to complain about how lengthy it’s taking display screen readers to deal with PDF, regardless that that drawback is behind us, let’s have a look at how properly display screen readers deal with HTML.

The fact is that HTML is a secure commonplace that display screen readers have had a very long time to get proper. (HTML 4.01 was revealed in 1999, XHTML 1.0 in 2000 [revised 2002], XHTML 1.1 in 2001.) However in actuality, display screen readers are nonetheless catching up. How is that this totally different from PDF help? It isn’t, besides in a method: It’s worse, as a result of HTML has been round longer. PDF help went from nothing to fairly good within the house of two years, whereas display screen readers are nonetheless moping alongside barely capable of deal with the total HTML spec.

In the event you’re making an attempt to recommend that the mixture of PDF-plus-screen-reader is an issue, what occurs if HTML-plus-screen-reader can be an issue? The grievance that display screen readers have bother with PDF and no bother with HTML is fake each methods. Why don’t we hear any complaints about having to improve display screen readers to deal with HTML?

Let’s have a look at the proof.

Model HTML help added
Model HTML help added
  1. headings
  2. longdesc
  3. accesskey
  4. onmouseover and onclick
  5. desk headers


  1. desk headers that comprise hyperlinks or graphics (whoops!)
  2. fieldset
  3. legend
  4. empty accesskey
  5. “ccented characters in English after they observe instantly after a hyperlink”


scope (and that’s apparently it)

Model HTML help added
Model HTML help added
4.0 (apparently)

  1. frames

  3. tables with headers

4.5 and later

  1. accesskey
  2. abbr, acronym
  3. title (and alt [sic]) on kind parts
  4. headings
  5. language codes
  6. lists
  7. longdesc
  8. object
  9. q, blockquote
  10. thead, tbody, tfoot whereas navigating by desk cells; th
  11. choose frames by doc title or body identify, title, or longdesc
  12. title on most, if not all, parts (additionally alt)
IBM Dwelling Web page Reader
Model HTML help added
Model HTML help added
  1. caption, headers, and abstract for tables
  2. alt for photographs and spaces
  3. title and longdesc for photographs
  4. noframes and noscript
  5. apparently additionally understands “paragraphs, desk cells/rows/columns, headings… lists, varieties, maps, and choose menus”
  6. “HPR helps HTML 4.0 together with widespread deprecated parts, however not SMIL, CSS, MathML, or DOM
  7. doesn’t help accesskey
  8. “tylesheets not supported on this launch” (!)


  1. label for choose
  2. optgroup
  3. legend for fieldset
  4. title attribute for abbr and acronym
  5. cite attribute for blockquote
  6. ruby (!)
  7. meta refresh
  8. Flash objects and the title for object
  9. body title and longdesc
  10. desk abstract
  11. disabled attribute for kind controls
  12. mouse and keyboard occasion handlers (akin to onmouseover, onclick, onkeypress)
  13. begin attribute for ol
  14. readonly attribute for enter and textarea
  15. accesskey
  16. tabindex
  17. maxlength attribute for enter
  18. “unsupported lang attributes [handled] by offering a setting to announce, however not converse, textual content which requires speech engines not accessible with HPR, akin to Greek”


  1. IBM and GW Micro have a behavior of ritually destroying launch notes for earlier variations when new ones come out. Right here, Window-Eyes launch notes had been gathered largely by Web Archive paperwork. A assessment of HPR 2.5 was used.
  2. Jaws 5.0 isn’t listed above. It was documented solely in rambling audio information (14 MB .exe). Apparently it added help for lists and blockquote (used for “indentation,” the recordings inform us).
  3. HTML help in Hal is tough to determine even after Dolphin Pc Entry despatched me the assorted launch notes. Model 6.51 fastened a Flash drawback and an issue with an offscreen-positioned skip-navigation hyperlink (marginally related to spec help); model 6.03 introduced, therefore acknowledged, hyperlinks, frames and headings, additionally abbr, acronym, and label.

So that you see, HTML help in display screen readers has developed and remains to be evolving. However impulsively when display screen readers needed to be upgraded to deal with PDF, some critics pretended that such upgrades had been unreasonable and distinctive. You’ve been upgrading your display screen readers all alongside simply to deal with HTML paperwork utilizing specs which can be as much as six years previous.

The place PDF accessibility falls down embarrassingly is with “authoring instruments,” the software program used to create PDFs. Only some applications can natively create a tagged PDF file, together with InDesign; PageMaker 7.0 (!); FrameMaker 6.0 and later; and Microsoft Workplace with an Adobe export plug-in (Workplace 2000 and later solely, Home windows solely). Merchandise that use PDFlib 6.0 and later can produce tagged PDFs. There could also be a number of different minor utilities right here and there.

The typical individual, nevertheless, shall be confronted with touching up an untagged or poorly-tagged unique. You just about haven’t any alternative however to make use of the tagging perform constructed into Adobe Acrobat (the total model, not simply the Reader, and for some capabilities you want the Professional model). There are already a number of not-very-helpful tutorials on tagging with Acrobat, and, on the threat of disappointing my readers, I’m not going to write down one other one, as life is simply too quick. Nevertheless, the fundamentals of what it’s a must to do are simple to state:


  1. Open your PDF.

  3. The Description pane of the Doc Properties display screen (File thực đơn) will inform you if the doc is tagged or not.

  5. If it isn’t, dismiss that display screen. Go to the Superior thực đơn and select Accessibility → Add Tags to Doc.

  7. Run a full accessibility test from that very same thực đơn.
  8. If the checker experiences any issues, open the little-known Tags palette (View → Navigation Tabs → Tags).Use the disclosure triangles to step by your doc’s new tag construction. You’re higher off if you choose Spotlight Content material from the palette’s Choices thực đơn, as Acrobat will then draw a hard-to-see border across the object whose tag you choose.

To deal with the commonest issues:


  • If Acrobat complains that your doc lacks a language specification, discover the topmost tag in your doc (instantly throughout the self-referential Tags tag). Proper- or Ctrl-click it and choose Properties. Choose a language from the pop-up thực đơn within the Language subject, or sort your individual two-letter language code.

  • For photographs missing a textual content equal, do one thing comparable, besides it’s a must to manually find the Determine factor that lacks the textual content equal. Context-click the Determine, choose Properties, and fill in Alternate Textual content (precisely like alt in HTML) or Precise Textual content (for an image of textual content).
    Properties screen

    (There’s a semi-automated method to discover all Figures with out textual content equivalents. In the event you run Superior → Accessibility → Full Verify and choose “alternate descriptions are offered,” Acrobat will discover all of the figures with out textual content equivalents and supply you hyperlinks to them in its report.)

  • A doc from a printed supply might comprise “artifacts” like headers and footers that you just by no means need screen-reader customers to listen to. You possibly can context-click on these objects (which can be deemed Determine, Half, P, or one thing else) and Create Artifact, which can trigger Acrobat and compliant display screen readers to disregard them when voicing. (You can even use the Contact-Up Studying Order instrument to pick the artifact on the precise web page and mark it as Background.)

If this job appears tedious, it’s, and it’s additionally fairly inaccessible to many individuals with disabilities. Keep in mind, we aren’t working towards an online wherein nondisabled individuals create content material for disabled individuals; individuals with disabilities should even be creators.

Acrobat is an uncommon program in that it should arguably comply each with the Authoring Instruments Accessibility Tips and with the Person Agent Accessibility Tips, as a result of you possibly can create and view content material utilizing Acrobat. (And PDFs themselves are topic to Net Content material Accessibility Tips; they are going to be lined in WCAG 2.0, which is predicted to be technology-neutral.) Acrobat and PDF are usually not totally compliant with any of these tips, however few issues are – and, with regards to ATAG, nothing is.

PDF accessibility isn’t as easy as HTML accessibility. However we have to stand as much as the untruths which can be spoken about PDF, particularly since a lot of these untruths come from authorities with the facility to search out authors responsible of discrimination.

PDF accessibility is OK a number of the time when it’s dealt with by competent authors with what few instruments can be found. All of these elements want enchancment, however let’s not faux we don’t have already got the facility to create accessible PDFs. We do.



  • Jacques Distler

  • Andy Dulson

  • Loretta Guarino Reid

  • Phill Jenkins

  • Greg Pisocky

  • Ted Padova

Leave a Comment