Accessibility

Motivations and Barriers

Summary of Motivations and Barriers

  • Accessibility is about making technology accessible to the full range of human experience, which includes disabilities.

  • General reasons why open-source software should be more accessible:

    • accessibility benefits everyone

    • access to information and technology as human right

    • enjoyment and creation for disabled people

  • Reasons very specific to open-source software:

    • achieve the ethos of open-source

    • solving root problems due to open-source’s scale

    • better represent needs of disabled users

  • Open-source software is currently not very accessible due to:

    • decentralized structure

    • uncertainty in implementation

    • different accessibility needs of open-source software

    • under-representation of disabled users

    • insufficient funding

Accessibility is about making the world accessible to the full range of human experience, which includes disabilities. For example, the physical world can be made more accessible for wheelchair users by having more buildings with ramps. Not only should the physical world be accessible, but the digital world as well.

More and more people are accessing the world digitally. Access to the internet is a human right. [citation for UN]. With access to the internet, people can access vital information, such as education [citation for various country laws to access education] or their banking information.

People with disabilities access the internet as well. Blind users can access text on websites with assistive technology such as a screen-reader, which verbally reads text out loud. Thus assistive technology empower disabled people to live autonomous lives.

An accessible world allows people with disabilities to not only access information, but also communicate and create. Helen Keller’s access to Braille, another form of assistive technology, allowed her to read about the world. She used Braille to communicate with others who were Deaf-Blind. It also allowed her to publish her experiences [citation for Story of My Life] and advocate for the Deaf-Blind community to the general public.

Accessibility benefits everyone. For example, captions for videos benefit not only Deaf viewers with complete hearing loss, but also viewers with partial hearing loss due to age. Captions also benefit viewers who cannot play audio out loud, such as in a quiet environment or if they have broken speakers. [citation to range of disabilities and situational disability]. Thus accessibility benefits not just “a few disabled people”[a] but all people.

These are general reasons why open-source software should be more accessible. However, there are reasons very specific to open-source software as well.

Open-source software should be more accessible to better achieve the aims and ethos of open-source. Open source prides itself as available to all, due to its XYZ license [citation]. Money and licensing should not be a barrier to creating art for self-expression [GIMP vs Photoshop] or creating mathematical graphs for research [matplotlib vs MatLab]. However, available does not necessarily mean accessible.

A project’s community meetings available for all contributors to attend, but inaccessible to some due to time-zone differences. For example, a community meeting held at UTC XYZ. For a contributor in New York City, this meeting would be easily accessible at their local time of 3pm. For a contributor in New Delhi, India, this meeting would be inaccessible at their local time of 1am. In turn, to make their meetings more accessible, some projects alternate their meeting times [citation, NumPy alternate IST-friendly meeting].

In a similar way, disabled developers may not be able to access documentation or navigate the user interface of open-source software. For a visually-disabled developer, an image can still be accessed via screen-reader. The screen-reader reads out-loud the image’s ‘alt-text’, which is a short description of the image. Images in documentation are valuable sources of information. However, images without ‘alt-text’ cannot be accessed this way, and the developer loses valuable information.

For open-source software that uses a graphical interface, a mouse can be used to interact with the elements in the interface, such as clicking a cell or hovering over a button. A developer with motor-disabilities may interact via keyboard instead. The developer can navigate across elements via the Tab key and select to interact with the Enter key. A user interface that is not keyboard-navigable cannot be accessed at all for the developer.

If a developer cannot access open-source software, they are much less likely to join open-source spaces. Contribution and community engagement is the lifeblood of open-source development. Inaccessible open-source software is less diverse, and diverse contributions make open-source stronger [citation].

Thus, to truly make open-source available to all, it needs to be accessible as well.

Accessible open-source affects not only disabled developers, but disabled users as well.

Open-source software should be more accessible due to a matter of scale. Open-source is foundational to modern digital infrastructure [citation… that xkcd comic?]. Problems at the foundational level are magnified at scale, and thus inaccessible open-source projects affect millions of users across the world [citation of chapter in book ‘The Making and Maintenance of Open Source Software’].

For example, the open-source software CodeMirror is used as part of the Firefox’s Developer Tools [citation]. CodeMirror provides a code editor in the browser. However, Codemirror 5 and its previous versions were inaccessible to keyboard and screen-reader users. It did not support the use of the Tab key as a navigational input. Instead, the Tab key could only be used for code indentation.

As Firefox is used by millions of users every month [citation on usage], millions of people have inherited this inaccessibility. Thus at scale, many users could not access Firefox’s Developer Tools. It was only years later with the release of Codemirror 6 that an option was provided to use the Tab key as navigational input as well. Rather than having projects built on top of open-source software inherit problems or create their individual patches, open-source projects have the power to solve accessibility problems across the board.

Barriers

The current state of open-source software is not very accessible, for a variety of reasons.

Open-source software is currently not very accessible due to its decentralized nature. Being decentralized is both open-source software’s greatest strength and weakness. It allows anyone to contribute, and different contributors contribute different things. However, this can also make it difficult to create a cohesive alternative experience for a screen-reader user, keyboard user and more. If an accessibility is not evenly applied, a user journey can be disrupted, such as with a keyboard trap [citation]. Thus to make open-source software more accessible, it must be done deliberately, in stages and in coordination with others[b].

Open-source software is currently not very accessible due to uncertainty in implementing accessibility. The vast majority of contributors in open-source software are software developers [citation on percentage]. While developers may feel comfortable contributing code, they may feel less comfortable contributing outside of the usual wheelhouse, such as accessibility. While there is growing awareness about accessibility in open-source [citation], contributors may still hold back due to a fear of ‘doing it wrong’. By providing guidelines in this document, we hope that this uncertainty can be overcome.

Open-source software is currently not very accessible due to the different accessibility needs of open-source software. Current accessibility guidelines are for general use and focused primarily on web-pages [citation, WCAG]. There are guidelines specific for certain sectors, such as education and the creation of accessible tests for disabled students [citation, Diagram Center and USA education]. However, these guidelines are not suited for open-source’s technical demands and processes. There are much less resources on how to make a command-line interface accessible, best practices in writing accessible documentation or community meeting notes. This document aims to provide guidelines specific for open-source software.

Open-source software is currently not very accessible due to the under-representation of disabled people in open-source spaces. While disabled people make up X [citation of disabled people in general population], in open source the number is Y [citation]. The above cited barriers prevent disabled people from using open-source software and reporting problems with their experiences, such as with Github Issues. These barriers also prevent disabled developers from contributing back. If disabled people are kept out of a space, then disabled people’s voices are not heard within that same space. It creates a chicken-and-egg problem where open-source software is not made with disabled users in mind and in turn disabled users cannot help shape open-source.

Open-source software is currently not very accessible due to insufficient funding for accessibility. Open-source software relies not only on volunteer labor of contributors, but also paid labor through grants [citation]. Grants allow proposed features to be prioritized and created in a sustainable way. However, out of X dollars spent on funding open-source software, only Y percent, Z dollars is spent on accessibility [citation]. This is related to the above mentioned problem of under-representation. In a grant proposal,it can be difficult to justify funds for making open-source accessible without proof of demand such as with a high number of Github issues about accessibility [citation of grant-proposal process[c]]. Thus this lack of funding perpetuates another cycle where accessibility is not financially supported.

Overcoming Barriers Together

Open-source software can begin to be more accessible by coordinating with each other. Within the scientific Python ecosystem, there is one such effort: the Scientific Python Ecosystem Coordination (SPEC).

The SPEC process is designed to identify areas of shared concern between projects in the scientific Python ecosystem and to produce collaboratively written, community adopted guidelines for addressing these areas. Such guidelines are known as ‘SPEC documents’ or a ‘SPEC’.

Projects in the ecosystem have an existing, diverse set of proposal processes and development constraints. SPECs complement these: they are a mechanism to encourage shared practices and improve uniformity of experience across projects. This SPEC captures established practices so that new projects can learn from them. As SPECs are living documents, authors may propose a new practice that they believe will benefit the ecosystem as a whole.

Projects decide for themselves whether to adopt any given SPEC—often, this would be through team consensus. A SPEC may not be a good fit for every single project, and thus there is no expectation that all SPECs must be adopted by all projects. That said, SPECs serve their purpose through being adopted by several projects—and their authority stems from the extent to which they are. [citation, SPEC Processes page]

The purpose of this SPEC document is to: - Make the ecosystem more accessible - Clarify the broad term accessibility and its various possible implementations - Provide specific assessment tools and steps - Direct to existing accessibility standards and modifications for open source specifically - Provide case studies of how different core projects became more accessible - Start discussion and develop shared solutions