Suggest a rootMargin strategy for lazy-loaded elements

[zcorpan]

See #5408 (comment)

Fixes #5408.

• At least two implementers are interested (and none opposed):
  • …
  • …
• Tests are written and can be reviewed and commented upon at:
  • N/A
• Implementation bugs are filed:
  • Chrome: …
  • Firefox: https://bugzilla.mozilla.org/show_bug.cgi?id=1618240
  • Safari: https://bugs.webkit.org/show_bug.cgi?id=203557

(See WHATWG Working Mode: Changes for more details.)

---

/urls-and-fetching.html ( diff )

[rwlbuis]

What is the rationale behind the value 1250px?

[zcorpan]

It's the smallest value that Chromium uses. They experimented with 750px but found it was too low. See #5408 (comment)

We could specify a different value, or not suggest any particular value. However, what I'd like to avoid is having some browsers use a very small value (like 1px or 50px) which makes the feature annoying for users and causes web developers to avoid using native lazy loading.

[zcorpan]

Maybe this could be goal-based instead, like saying there should be a minimum value that most often results in images being loaded before they intersect the viewport under normal usage patterns for the given device. This could be quite different for a watch compared to a smartphone, for example (because smaller screen and maybe slower scrolling).

[rwlbuis]

Right, I was thinking this should be carefully worded as we can have many different form factors (watch is a good "extreme" case).

[zcorpan]

OK, please see 88384f6 & a9f9c1d

[zcorpan]

@scott-little @addyosmani @domfarolino @smfr @emilio

[scott-little]

LGTM - I like the wording here. I agree that having a goal-based root margin description is probably best here.

[addyosmani]

Maybe this could be goal-based instead, like saying there should be a minimum value that most often results in images being loaded before they intersect the viewport under normal usage patterns for the given device.

I concur with this approach of being goal-based in thinking about how to approach the rootMargin strategy too. LGTM.

[domenic]

Editorially LGTM. I'm not sure whether we should wait for more review (e.g. from other implementers), or go ahead and merge, since it's just adding suggestions. I'll leave that up to you, @zcorpan.

[zcorpan]

Thanks, @scott-little and @addyosmani!

@domenic, let's give this a day or two for review from others. I can merge on Thursday if there are no objections or new feedback.