compare domvm with mithril, performance etc.

I’m just found domvm, and used mithril for a long time before, now I’m doing a performance test as below, using

https://github.com/MithrilJS/mithril.js/tree/next/performance

And your mithril-adaptor

DOMVM2.1.4

rerender without changes x 358 ops/sec ±10.57% (29 runs sampled)
construct large VDOM tree x 11,796 ops/sec ±5.39% (32 runs sampled)
mutate styles/properties x 287 ops/sec ±20.81% (21 runs sampled)

Mithril 1.1.1

rerender without changes x 3,414,552 ops/sec ±6.55% (47 runs sampled)
construct large VDOM tree x 3,610 ops/sec ±4.56% (49 runs sampled)
mutate styles/properties x 10,763 ops/sec ±3.24% (51 runs sampled)

Seems the construct time is fast, but later mutate styles/properties is slow for DOMVM

But I’m wondering some lost in my test, can you have a look at the test file for DOMVM?

test-perfdomvm.zip

[leeoniya]

hey @futurist domvm focuses on doing one thing: rendering and composing views.

domvm-router does not enable defining routes in your views, it’s simply a router i wrote, similar to existing third party routers as mentioned in the docs [1]. domvm-router does have a nice feature of generating routes from parameters, which many other routers lack. it does need some documentation, but you can see how it’s used here [2] and feel free to use any other better-documented router lib instead.

the logic of which views must to be rendered based on route can be defined in your templates, it can be as simple as:

function activeTabView() {
  if (window.location.href.indexOf("users") != -1)
    return vw(UsersView, store.users);
}

el(".tab-content", [
  activeTabView()
])

[1] https://github.com/leeoniya/domvm#what-domvm-is-not [2] https://github.com/leeoniya/domvm-router/issues/1

[iamjohnlong]

@futurist https://rawgit.com/krausest/js-framework-benchmark/master/webdriver-ts-results/table.html