Frequently Asked Questions
What is ObjectDescription?
ObjectDescription is a service that lets anyone write a description for any non-textual object (such as an image) on the Web.
Who can use ObjectDescription?
ObjectDescription is for all users of the Web. Anyone can write or edit a description for an image. In the future, anyone will be able to search through image descriptions for valuable information that is inaccessible today.
Are there any fees to using ObjectDescription?
No. ObjectDescription is free.
Why is it a good idea to describe images?
Images are an integral part of the Web. To make the Web more usable for everyone, we need to make images accessible through textual descriptions that can be processed by people and machines.
How will textual descriptions of images benefit everyone?
- Image descriptions are treated as standalone documents. This means they can be processed by machines such as search engines, and can be included in regular search results. This can increase the amount of searchable data on the Web and yield more relevant search results.
- Image descriptions can make image searching a lot more efficient.
- People speaking different languages can describe the same image. Since image descriptions can be made available to automatic translation tools, descriptions can be translated and made available to more people speaking different languages.
- People can use descriptions to get more details about images. For example, a description of an image of a graph can be a table, with precise data that can be copied to a spreadsheet.
- People who are not able to see images can instead listen to image descriptions, using assistive technologies. This can include those with limited or no vision. In the future, it will include browsing the Web using hands-and-eyes-free browsing technologies such as in-dash car computers or Walkman-like products.
Doesn't alternate text in HTML already make images accessible?
No, the function of alternate text is not to make images accessible. The role of alternate text is to make content that contains images comprehensible when images cannot be seen. What's the difference? The difference is that if alternate text makes images accessible, then reading alternate text for a missing image should sufficiently describe the image. While this may work for some images, it does not work for all.
For example, as seen below, reading just the alternate text "love" in place of the heart image works as alternate text. If the image is not viewable, the sentence will read "I love you". But "love" does not adequately describe the image itself:
<p>I <img src="heart.gif" alt="love" /> you!</p>
The key is to remember that the function of alternate text is to work with surrounding content to make sure that content remains comprehensible (and therefore accessible) when images cannot be seen. Image descriptions are descriptions of what is in the image and are not affected by surrounding content.
Does the longdesc feature in HTML make images accessible?
Why do we need a replacement for the longdesc feature?
The longdesc feature can be useful if authors use it. But getting authors to use it has been hard primarily because of the shortcomings of longdesc.
ObjectDescription lets authors continue to use longdesc now. And the content generated now for longdesc will continue to live on and be useful in the future since the ObjectDescription API will seamlessly replace the need for longdesc.
What are the shortcomings of longdesc?
- The longdesc feature has the unfortunate name "Long Description". The name incorrectly suggests to authors that alternate text is a short description, while longdesc is just a longer version of alternate text. Authors don't see a need to write two descriptions that are only different in length.
- The longdesc attribute must be physically added to an image reference in the document that uses the image. The person who may want to describe that image may not be the author of the document and therefore may not have access to the document.
- Authoring tools (WYSIWYG editors, CMS, etc.) don't make it easy to create the documents that the URLs for longdesc point to. ObjectDescription makes these documents readily available for everyone.
Is the future HTML feature called "aria-describedby" a suitable alternative to longdesc?
Unfortunately, aria-describedby only makes images accessible under limited conditions. It's a poor alternative for the longdesc feature because:
- It can only point to an image description within the same document.
- Even though it can point to structured content such as bulleted lists or tables, aria-describedby can only process structured content as plain text. This is because the accessibility APIs that aria-describedby maps to do not support structured content.
- Authoring tool vendors are likely to find it difficult to build user-friendly interfaces that use the aria-describedby feature. This is likely to result in confusion and little use of the feature.
Why is the ObjectDescription API a good replacement for longdesc?
APIs are methods that computers use to talk to each other without users being exposed to the behind-the-scene mechanics. Thus, the ObjectDescription service API will enable applications such as Web browsers to give authors the ability to create and view descriptions of images.
For example, in a future browser that uses the ObjectDescription API, users will be able to get a description of an image in the Image Properties dialog box.
And anyone will be able to write/edit an image description directly from the browser:
Why is it important to let anyone write image descriptions?
The task of making the Web accessible is too big to be accomplished by writers alone. Other users of the Web must be permitted to get involved to make the Web useful to everyone.
Will permitting anyone to describe images be abused?
All open, community-led efforts are exposed to abuse. Abuse can take the form of spam, mischief or poor quality data. The goal of ObjectDescription is to be useful and this can be achieved while tolerating a small amount of abuse. But, automated and human efforts will be taken to curtail abuse.
How does ObjectDescription work?
When used with longdesc:
- Users enter a URL to an image and write a description for that image.
- ObjectDescription gives the user a URL to enter into the longdesc attribute.
When used with the API:
- An application like a browser takes a fingerprint of an image and submits it to ObjectDescription.
- ObjectDescription returns the description of the image.
- If the user edits that description in the browser, the description and the fingerprint is sent to the ObjectDescription service for storage.
What is a fingerprint of an image?
A fingerprint is a unique number that is generated by examining every bit of an image. Even a tiny change to an image will create a different fingerprint number. There are many different methods to generate a fingerprint but ObjectDescription uses a method called MD5. Here is what a fingerprint looks like:
Does ObjectDescription store images on its servers?
No. Only descriptions of images.
Will there be support for non-English image descriptions?
Yes. Soon. In addition, this website will be available in other languages in the near future.
Can I instruct future browsers to use a different object description service?
Yes. Add the following HTML to a webpage:
<meta name="objdesc.service" content="... URL to another service ..." />
Can I instruct future browsers not to access descriptions for images on my website?
Yes. Add the following HTML to a webpage:
<meta name="objdesc.active" content="false" />
Can I write an image description that others cannot modify?
Yes, This feature is coming soon.
Who owns user-contributed data (i.e.: descriptions)?
We claim no rights over user-contributed data, such as descriptions.