[BUGFIX release] Deprecate {{unbound block and multi param usage.

[rwjblue]

This directly targets the stable branch, as this functionality has been removed from beta and canary already.

[workmanw]

So to be clear, is that this code snippet?

{{#unbound if isActive}}
  {{!-- Stuff --}}
{{/unbound}}

Is there an alternative? We actually got a good bit of performance from {{unbound}} because we had a lot of complicated read-only data.

[rwjblue]

@workmanw - Yes, exactly that type of usage is what is being deprecated. You should refactor to nested helper usage. Something like:

{{#if (unbound isActive)}}
  {{!-- Stuff --}}
{{/if}}

[workmanw]

@rwjblue Excellent. Thank you.

I've been seeing this helper/subexpression pattern more and more. Is there anywhere I can learn more about these and other changes? I fear that I would have missed this if I hadn't looked through the change log.

[rwjblue]

I've been seeing this helper/subexpression pattern more and more. Is there anywhere I can learn more about these and other changes?

Hmm, no. I am not aware of any specific articles or posts that are focused on nested helper (aka sub-expression) usage.

I fear that I would have missed this if I hadn't looked through the change log.

Awesome, I am glad the changelogs are getting more useful (since I started linking PR's)!

[workmanw]

Thanks for being awesome!

[mmun]

Don't use unbound for performance. It is slower in Glimmer and eventually will be removed.

[workmanw]

That's a bummer. It's really help us stave off performance issues.

Some background: we have several views that are read-only and based on polymorphic types. In one of our more complex views, we show up to 250 of these components at once, each component has 10 different possible renderings we used unbound because the type can never change and it reduces the binding/setup on the {{if}}s.

Eg:

{{#unbound if isTypeX}}
{{/unbound}}

{{#unbound if isTypeY}}
{{/unbound}}

// 8 more of these

So we found unbound cut the rendering time by like 35% for these components. Any suggestion on how to handle this in glimmer? There isn't much out of the box initial rendering performance yet that we've seen.