Skip to content

Hackday #20, Jan 26, 2022

Jump to bottom
Manu edited this page Jan 31, 2022 · 4 revisions

Attending

  • Manu Heim, ETH Zürich (moderates the morning)
  • Josua Muheim, Nothing (moderates the afternoon)
  • Luis Argüello, ETH Zürich (in the afternoon)
  • Esther Brunner, Zeix
  • Miro Dietiker, md systems
  • Dragan Nikolic, Zeix
  • Petra Ritter, Access for all (only in the beginning)
  • Donato Rotunno, Liip (in the morning)
  • Stephanie Rotzetter, Iterativ
  • Jürgen Rudigier, Dept (eh. "Hinderling-Volkart")
  • Graciela Schütz, Unic
  • Oriol Torrent, Unic
  • Ramon Wenger, Iterativ

Timetable

  • 09.15 Welcome and housekeeping
  • 09.30 Re-boot our mission:
  • 10.15 Accessibility Developer Browser Extension "Epiphany": Presentation by Josua, Feedback by all of us
  • 12.00 Lunch break
  • 13.15 Accessible multi-selects: currently, there doesn't exist a fully satisfying solution. Should we do a pilot project together and create one?
  • 15.00 End

Welcome and housekeeping

Major changes at Access for all. Manu left A4A last summer. Josua has given up his percentage at A4A, as well. By the end of December 2021 also Franco D'Elia and Sylvia Winkelmann-Ackermann have left A4A. There is a new director called Philipp Keller but the ADG will most probably not be on his priority list right now.

We will have to find out how to continue with the our powerful group that has gathered around the ADG: is our old mission still meaningful? Or should we re-align to find a new mission? Josua and Manu have talked about it and do have some ideas for the future. We would like to hear your ideas, too.

  • Esther: We should still keep this group but we need a broader vision, sharing accessibility knowledge in CH. The ADG will still be the main tool where you document our best practices. The biggest plus in this group is that there is so much knowhow in this group.
  • Stéphanie and Ramon: The group is so useful because there is so much knowledge and we can learn from each other, it is a good way of staying in contact.
  • Jürgen: Let's keep the ADG up as it is. We might have to rework the pages to make it more a reference/guide than a book. That could maybe make it even better to read. Agree on the group, that it should rather be community than the group that works on the ADG.
  • Josua: Love the spirit of having this group, of the working in that same spirit. People are still around, even though we could not meet in person. There won't be big solutions like Epiphany, keep it low-level.
  • Miro: A group would be helpful. With the ADG, if there is no attribution to those people that work on the ADG, it is not attractive. Josua: At least, add a page with those names of the persons that work on it. And maybe add a link to meet-up and say that the discussion goes on there.

Morning topic: Epiphany

Josua presents Epiphany, slides can be found at https://slides.com/nothinginc/epiphany/. Nothing AG sees itself as a Venture Lab rather than an Agency (Nothing): Creating products. Creating ideas. Josua created epiphany and he shows it to the group.

About Epiphany:

  • A digital mentor, like Karl Klammer (Clippy) in Microsoft Word.
  • What technical invention would render your job obsolete?
  • Anna-Karenina-Principle translated to accessibility: "Alle hochwertigen Websites gleichen einander; jede mangelhafte Website ist auf ihre ganz eigene Art mangelhaft." (Josua Muheim, 2021)
  • Clippy was not really helpful. But imagine a friendly virtual mentor watching over your shoulder while being at work ** designing a GUI ** implementing some components in HTML ** writing content or producing a video ** etc.

Reactions:

  • Donato: Likes the idea. Question: Did you think about automatisation, as well? Josua: If we start on new project, blank page, then it would probably work to integrate Epiphany into automatic testing tools. Not separately, though. This one-solution-for-one-use-case principle is really important in Josua's opinion.
  • Miro: Doubting that developers will re-apply their knowledge once they have learned it. Why should they, if they have not listened to us accessibility consultants. Plus there are so many tools already, it won't work to build up another single tool, where adoption will be hard. Josua: Not aiming to big players. Josua sees Epiphany as a solution that would help smaller players to get better and to get them into the movement. Epiphany would not be there for big projects. Josua thinks their would be an audience for a tool like Epiphany. Miro: His company tries to sell services, e.g. how to implement Drupal correctly. Some companies do adopt this culture, but many don't.
  • Ramon: Really likes the idea. Ramon never liked Clippy, so he works in his own tools. So he would like such an assistance, but he'd like to have them in his own tools. There seem to be some tools that can be integrated. Josua: There are helpful tools, command line tools, he is not very much into that yet. Ramon: In his team they all try to write the same code. But in the end all have their own way of coding. There are so many rules that would be taken into account, he'll sure forget about some of them. Josua: Automatisation, Pipelineability, the system Epiphany should be able to use from different sides. Ramon: Needs something to feel forced/motivated to really do accessibility.
  • Esther: Where is the right place for this mentor to live? Where should it be, in a browser, in the developers' tools, etc.? Esther feels it would have to be in the dev. tools. But that might be hard, though. Rather in a browser? But it feels a bit like re-inventing the wheel. In the end it will be a tool like Siteimprove and that's a big competitor. Plus: The simple cases you presented can be handled well by the existing tools. The difficulty lies in the complex cases, and that's where automated tools fail now. The solution would have to be better than these already existing tools for complex cases. Josua: ((I did not know what to write here)) Dragan: Great idea, shares Josua's notions that it is frustrating to be asked late for accessibility. Further reaction: While coding, not everything will be on one page, rules won't detect many of the ideas that the developers work on. It would be more of a discipline thing for developers to have accessibility included in every sprint. You can e.g. use Lighthouse in your development sessions. They work well for all those things that you can easily measure. There's a google Chrome browser in a headless environment, that can be helpful. Probably around 80% of the accessibility cases can be solved easily, but then there remain the other 20% of cases where you will find an individual solution every time. Josua: Still thinking that there should be a chance to do it correctly from scratch. Dragan: Lots of frameworks, makes things harder. And many times one works on atoms' level. Josua: Maybe in the far future frameworks will rely on those helpful solutions.
  • Donato: The idea is really good, and there would have to be some finetuning. Josua: All those tools, they try to take into account all those different ways how you can do stuff (in coding in general but also in a11y). Epiphany would just propose one (perfect) solution. Kind of a best-practice mentor.
  • Miro: Why not strip ADG from those unnecessary solutions? And bring in this one-perfect-solution into the ADG. Josua: THe problem is that the ADG is still pull, they have to go out and get it. Miro: Maybe just add it to the ADG.
  • Graciela: Interesting idea Josua... I have to go, see you in the afternoon :-)
  • Miro: Maybe the AGD needs a section / excpansion about (current) QA tools ;-)

Afternoon topic: The quest for a truly accessible multi select

The ETH Zürich needs a reusable filterable multi select component

  • Previous investigations have not yielded a top candidate so far
  • The ETH would probably offer some funds to find a solution to this problem
  • The question is: is this an opportunity for the ADG team to develop a solution?

We (re-) evaluated a few multi selects, incl. two overlooked candidates:

In accordance with Luis (who works for ETH), we see a short-term and a long-germ goal:

  • We need a short-term solution which works "well enough"

  • A reusable proof of concept would work for this (similar to the ones in ADG → Examples → Widgets)

  • We could build upon a text input to filter either a or a list of checkboxes ( was mentioned, too, but its accessibility does not seem to be supported enough)

  • It does not need to be a full-fledged solution, but it needs some advanced features like loading options using AJAX

  • It also needs to be mobile-friendly: a good idea is to just offer a native there, as they have a really useful native GUI on mobile (in sharp contrast to the ones in desktop browsers)

  • Luis hopes to have a solution like this in 2 months - but this is very "sporty"

  • We need a long-term solution that is really good (and would replace the short-term solution)

  • This could either be an existing library that we help to optimise its accessibility (taking into account the learnings from creating the proof of concept)

  • Or it could be creating our own one (based on our proof of concept), although there's consensus that this means a lot of work (incl. maintenance)

In general, all attending members feel that it would be an interesting opportunity to try to tackle this situation together

  • Everybody could bring their expertise into the mix: Josua delivers the basic idea and how it should probably be done, then developers implement it together iteratively (while Josua regularly tests the results)
  • Even members who don't see a specific involvement can be very helpful in supporting the organisation of the tasks, meetings, etc.
  • We all would definitely learn a lot from this, and it could bring us closer together again under a clear mission
  • We have to be aware though that we cannot guarantee that this "field test" will work out in the end: there might be reasons why nobody has tackled this perfectly so far
  • On the other hand: if we manage to tackle it, this would probably have the potential to create quite some excitement in the a11y scenes around the world, and could foster attention towards our ADG (and/or whatever new mission we take on)

To proceed with this, we first need to define clear requirements that ETH has for the "short-term solution", and what financial amount they are willing to offer. @Manu and Luis, please do so!

  • From there on, we can become more specific and talk about the details.
  • Notice: Jürgen mentioned that 2 years ago, the ADG team decided not to accept any money! We certainly would have to talk about this again as a team.

Next Hackday

Manu will send out a doodle for March/April.

Clone this wiki locally