Canvas MDN Short Survey proposal

[atopal]

This is a proposal for an MDN Short Survey for Canvas, based on the State of Graphics discussions.

Posted under https://developer.mozilla.org/en-US/docs/Web/API/Canvas_API and other associated canvas documentation, possibly also SVG

Survey Question 1

Purpose: Understanding the distribution of usage so that we know what areas to focus our investment on.

Question: What category would you say your use of canvas falls into? Check all that apply [checkbox options] :
Answer options:

1. Playing custom animations like Lottie or Rive
2. Charting & graphical information
3. Games
4. Creating an editor (ie photo editor or design tool)
5. Showing 3D models (outside of games)
6. Art
7. Other - please specify [short form field]
8. I’m not using canvas

Survey Question 2

Purpose: To get a signal of what the biggest weakness within the graphics space is based on the question 1

Question: How would you rate the ease of use of canvas for your use case? [Scale of 1-5]
Answer options: For each of the selected areas above

• Very difficult
• Difficult
• Neutral
• Easy
• Very Easy
  Recommend adding: “I’m not using Canvas” to all fields

Survey Question 3

Purpose: trying to dig into some specific priority questions we have in each of these areas

Use case category specific questions – will be shown if and only if Animations (answer 1 to question 1) is selected

Question a1: What animation format do you use[checkbox options]
Answer options:

• SVG
• Lottie
• Rive
• Animated PNG
• other [form field]

Question a2: Are there animations or specific animation features you cannot use effectively due to lack of performance or other reason (please specify)?
Answer options:

• [free form, short]

Question a3: If all browsers supported the Lottie format natively (no need to ship a JS player), would you invest the time to stop shipping the player approach and rely on browser support instead?
Answer options:

• Yes
• No
• Unsure

Question a4: if just one supported the Lottie format natively (no need to ship a JS player), would you invest the time to stop shipping the player approach for just that browser?
Answer options:

• Yes
• No
• Unsure

[tidoust]

Nice. Questions 1 and 2 seem good to me. Regarding question 3:

• a1: The list sort of surprises me as the first formats that come to my mind for short animations are animated GIF and PNG. That's the first ones I would attempt to use. Am I too old already?
• a3: The question seems to call for a "Yes" answer. Would you stop relying on a polyfill if it's not needed? Well, yes! Why would people choose "No"? I suspect you're trying to evaluate something else. What did I miss?

[atopal]

Re a1: I think this is more about animating things as part of a site, not dropping objects into the page that have been animated somewhere else

a3: yes, this might require more nuance as a free form response or adding the trade-offs into the question.

@ThomasTheDane can you comment on the above?

[ThomasTheDane]

Sorry for the delay -

• a1: great point I have added these two options to the doc version of this. Also Lottie and Rive should be different options which I think they are in the doc but not this post.
• a3: True - we really want to know if they'd go through the effort to update it and switch it out. I updated it to " would you invest the time to stop shipping the player approach and rely on browser support instead? ".
• I added an a4 question as well: "Question a4: if just one supported the Lottie format natively (no need to ship a JS player), would you invest the time to stop shipping the player approach for just that browser? "

[atopal]

Updated the proposed survey

[jgraham]

If all browsers supported the Lottie format natively

Asking about specific native support for a proprietary format is very surprising to me, and I'm not sure what the thinking behind the question is. Is there actually a credible plan to ship this format inside browsers, and what does it look like?

For a question coming from a standards group, I'd assume that the proposal is for a web-standard that's compatible with the existing format. If that's the case, is the existing discussion somewhere? Are the people making Lottie on board? I can imagine them being very surprised if browser vendors want to turn their proprietary product into a royalty-free standard.

If the thinking isn't a web standard then what is it? That some proprietary browsers would do a deal with Lottie to bundle an implementation, and that other browsers would continue to use the existing web implementation? That seems extremely problematic.

I don't think this survey should run with the questions about including Lottie directly in browsers, without clear confirmation from the people working on it that they're on board with this kind of question, and confirmation that any work to implement it would be as part of a web standard.

[tidoust]

Asking about specific native support for a proprietary format is very surprising to me

I'm not familiar with the Lottie format but I see there is a Lottie Animation Format spec, which seems to come with some open governance model and RF commitments from contributors. The community effort seems to have started in 2024. Am I not looking at the right thing? That may not be formal standardization, but format seems open at least?

[jgraham]

Interesting, I hadn't found that; thanks!

I would still be very reluctant to send out a survey asking if that format should land in browsers without explicit buy-in from the people working on that specification for that as a possible plan. Does such buy-in exist?

[ThomasTheDane]

Product manager of Chrome graphics team here to verify that yes, we have buy in. My team, specifically Florin Malita, has been driving the spec and he is starting the work to land Lottie support directly in Chrome as a new video codec format. On the Chrome side we have decided we'd like to support Lottie and this question is meant to help us understand how critical cross browser support is to developers. As noted there is a formal spec and we would be just shipping what is in the spec so no proprietary concerns.

[jan-ivar]

Hi @ThomasTheDane can you clarify what you mean by "video codec"? I thought Lottie was a JSON-based vector animation file format.

[ThomasTheDane]

Lottie is a JSON based format in how it is described and we're exploring the best mechanism by which we'd expose it through the web platform as a "playable" file. Making it work as an image format similar to gif was considered but that lacks seeking support (can only start and stop) which is a deal breaker. Introducing it as a video codec that is playable and controllable explicitly seems like an appealing option. We're happy to hear more input on the best way to implement it at a browser level but I think that's orthogonal to the questions we're asking in this survey.