Package name

[dylanb]

Thank you for contributing to the axe-core® and axe® community

We are little concerned about the licensing related to the algorithm that you have implemented in this package. For this reason, until these concerns are addressed, we would prefer if you renamed this package to something that did not contain the name axe in it.

We were thinking apca-w3 or color-contrast-apca, but of course the choice is yours

In either case - please note that you should use the Deque registered trademarks in accordance with the Deque trademark us policy

[meduzen]

Hi, not a maintainer here, but:

1. I’m curious to know which specifics led to this post. As a potential user of it (in a private or professional context), what is concerning about the licensing and/or its related code?
2. Maybe it’s in the best interest of Deque to provide guidelines making it easy for the community to build and discover ways to enrich their use of Axe (e.g. by recommending ways to name packages, like “all Axe core 3rd-party plugins can’t have this naming pattern and should use that other naming pattern: e.g. use axe-core-plugin-apca or axe-core-rule-apca instead of axe-apca”). Changing the name of the package for something less obvious (like apca-w3 or contrast-color-apca) would likely not meet this goal. Also note apca-w3 is obviously not available as it’s the main dependency of this project.

[acorncom]

One potential way of handling the "can't find plugins that tie into axe" issue could be to use keywords in package.json. That would allow keywords like axe-core-plugin or axe-core-rule to be specified (assuming the axe team is happy with that approach) without the actual package name falling afoul of trademark policies.

@dylanb would that type of approach handle the concerns on your end while still allowing the community to find plugins for axe-core?

[Myndex]

Please allow me to address the license issue first:

The current APCA license is restrictive because of the unfortunate situation with the “cult of obstructors” who were at one point creating bad unauthorized implementations to create a bad impression of the work we are doing.

It is otherwise permissive, and once we get into final it will shift to a permissive model.

Next

I am well aware that Deque is opposed to changing the existing (poorly functional) WCAG2 contrast math.

It took me a while to understand why, but I think it’s because (correct me if wrong).

1) the Deque business model surrounds automated testing, and

2) a certain person planted a “bad seed” claiming that APCA would not fit an automated paradigm, which is utterly false.

Maybe I’m wrong, but just in case I’ll take this opportunity to point out that unattended automation of color/contrast is a principal goal of perceptual uniformity. Not only is APCA automatable, the fact is you cannot have useful automation without a perceptually uniform solution.

If Deque wants to discuss, explore, or see what it’s all about, I’m open for a zoom call next week.

Thank you for reading

[giamir]

@dylanb As requested we have now renamed this package to apca-check. We have also added the following statement to the license as specified in your company trademark policy: axe-core® and axe® are a trademark of Deque Systems, Inc. in the US and other countries.

I do agree with @meduzen that would be beneficial for the axe® community to have clear contribution guidelines/conventions about how 3rd party packages for axe-core® should be named.

I am closing this issue since it is technically solved but feel free to continue discussing any open point in the thread. Thank you.

[dylanb]

@giamir thank you for your flexibility. We are in the process of communicating our trademark policies to the community in general.

Specifically as it relates to APCA - we do not want to have any code included or promoted as "open source" or "free" while there are serious potential intellectual property concerns that could end up significantly negatively impacting the adopters of tools that are based on it.

I personally find it problematic that the license for this repository is listed as MIT, when these issues are known

[Myndex]

Re APCA: I have mentioned this is a non issue, and we have provided permissive licenses (AGPL 3) on request. So long as the code is implemented correctly.

There is a small cult of obstructionists who have attempted to create discord with incorrect implementations, and that is the only reason for the license as stated, so that we can protect the development.

Once the official recommendation is set, the general license will shift to a permissive model.

Thank you for understanding.