New polyfill

[fvsch]

So we have a first polyfill in this repo, and a different approach at:
http://codepen.io/jonneal/pen/EKPONK
which I expanded upon at:
http://codepen.io/fvsch/pen/MyyBXW

So whether we make a fitie 2.0 or a new repo with a different name, it would be good to have a broad idea of what we want to do.

Name

"fitie" might be problematic with Edge and older WebKits needing some polyfill too.
If we don't keep this name, I guess we need a new repo.

Feature set

Rationale for some of the changes I did in my codepen:

• Add (limited) object-position support: because it's nice. :) But it could be left to users, with instructions to set both object-position and background-position.
• Move background-size and background-position out of the JS and to the CSS: I think it's nice to have the JS only add the background-image, and let users control it in CSS. Likewise, it's probably better to avoid resetting background-color.
• Expose a global function: allows users to reapply the polyfill to new content (has to be done manually); there's some code to make sure we don't work on the same image twice.
• Add limited responsive image support: using currentSrc then removing srcset.

Do we want to keep all of that? Only specific bits?

[fvsch]

Looking at fitie.js, you're wrapping the img or video element in a custom element (duplicating a lot of CSS styles from that element to the wrapper), and then you have a little bit of complexity to support img and video and fill, contain and cover.

I guess that's needed for supporting <video>, but can you confirm that for <img> only we can go with background-image and avoid the complexity of doing the math (and live updates) ourselves?

[fregante]

Most of what you suggested has been published here: https://github.com/bfred-it/object-fit-images

• browser does the math through background-image
• object-position support
• background-color is not overridden (2.2.0)
• global function
• automatic new content detection
• srcset support

fitie's disadvantages are necessary to support <video> so it's unlikely that they are implemented here.

[DanielRuf]

I recommend archiving this project. Other solutions have better support and it is irritating when there are multiple polyfills (come on devs, GitHub is actually for collaboration ;-) ) and you have to decide which one to use.

[jonathantneal]

I’m okay with someone contributing to this project (with a non-failing PR) and eventually taking it over to make such a decision. That being said, bfred-it wrote:

fitie's disadvantages are necessary to support

Unless you have evidence to show this is no longer the case, I’m not sure what you are basing your opinion on, other than your personal convenience and current non-video needs.