Descriptive text for all images in a Blorb file

Following up on the “cover art text description” thread, it has been suggested that all images in a Blorb file could have textual descriptions. And sounds, too, because why not.

Proposal for an addition to the Blorb spec:

For a game to be fully accessible to visually impaired users, it should provide textual descriptions to be displayed as alternatives to images. Similarly, audio resources should have textual descriptions as alternatives for hearing-impaired users.

The resource description chunk allows the Blorb file to offer this information.

4 bytes         'RDes'          chunk ID
4 bytes         len             chunk length
4 bytes         num             number of entries
                ...             entries

The entries are variable-length, and look like:

4 bytes         usage           resource usage
4 bytes         number          number of resource
4 bytes         length          length of text (bytes)
length bytes    text            textual description (UTF-8, not null-terminated)

There should be at most one entry for each resource – that is, each (usage, number) pair.

Resource descriptions are not required, but they are recommended for significant sounds and images. (Images used for decoration, such as window borders or text dividers, may not need textual descriptions.) Data and executable chunks do not need descriptions; if they appear in this chunk, the interpreter can ignore them.

(An interpreter with a web interface would apply the textual description of an image as an “alt” attribute on the tag.)

What a very good idea.

Will the next version of Inform 7 have an option to comply with this proposal?

I believe so, yes.

Something like

Figure of Woodlands is the file "Woodlands.png" ("A dense forest with a few sunbeams.").

I’ve posted this change to the Blorb spec (eblong.com/zarf/blorb/). I’ve also updated my blorbtool.py script (same web page) to parse and display the resource description chunk, if one is found.

One question. Does every image gets a optional accompaning sound clip or does every sound clip gets a optional text description?

Every sound clip can have an optional text description.

The other idea is interesting, but better handled by the interpreter (using text-to-speech on the text description).

I was thinking more on the lines of displayinga image of a haunted mansion and playing a spooky midi simultaneuosly by ‘attaching’ the two together but I guess it can be done even now without the ‘attaching’ part. Thanks.