Interesting! Could you join a resource that list the superset of functionality. I only know that you can then nest other elements inside it. In "Form Design Pattern" it is stated that this variant have the side effect in older version of IE to submit the value of all the buttons to the server, regardless of which one was clicked, so you'll need to know which button was clicked so you can determine the right course of action to take. Maybe this is worth mentioning.

Hi @morgaan - I just offered to write this so we could rid of an issue that has been open for two years. To be honest, I'm not familiar with the implications, I was just summarising a conversation that we had in Slack years ago.

If you're more familiar with this than me, feel free to amend the text that I've submitted!

Hi @josebolos - First of all apologies that I did not mentioned it was an existing issue pending for long time. I thought it was coming out from work you being busy with lately.

I do not know much more TBH, I just remembered these lines from the book and I have been using mostly the input variant over the years, unless I had the need to nest stuffs in there. This is what is maybe meant by superset of functionality.

IMO I would let it open to developers to pick from, depending on their use case.
Reading the "We don't do this" looks like the input variant is bad practice which I doubt it is, at least according to Adam Silver, author of the book.
This may though be worth mentioning that in case of using the button variant that it does not fall into the edge cases category.

What do you think? We could also bring this topic into the next FE open space to see what is the fellowship opinion on that matter 2 years later.

In any case thank you for offering to write this ☀️. I plan to come up with a draft of the few tips I got from the book at writing forms, and already applied to NatCar project. This one could end up being a guideline chapter into the Playbook about writing forms and or a part in the Elements design system styleguide. It could also include the outcome of the current discussion.

Cheers!

@morgaan sounds great, thanks very much!
I also vaguely remember @davidpauljunior having some thoughts on button vs input, do you have some David? :)

older version of IE

Do we know which older version of IE? If it's old enough that it can't support TLS 1.2 then it won't be clicking any buttons on any of our sites because it won't be able to access them at all: https://github.com/springernature/frontend-playbook/blob/master/practices/graded-browser-support.md#browsers-without-tls-12-support

Still in IE11 as far as I understand. But it also seems that you should be safe if you provide the button with name and value attributes. In such case we should emphase this in the Playbook if using the button variant.

We definitely don't do that, since <input> is a self-closing tag. 😉

Indeed, well spotted 🏅