Sultan

Yes, this is not for the bench, just a proof of concept of a compiler that generates this format from a string source, related issue [#16](https://github.com/cssinjs/standard/issues/16)

For the tests to be fair you'd need to exclude the creation of the arrays from the bench and put that into the setup phase otherwise you're also measuring the...

Since the different methods use very different data-structures that are created on each run it would not be comparable since the creating of the data-structures is meant to be a...

You're also measuring the cost of creating the function on every operation i.e `return (input) => {` this could throw out any possibility that the VM could generate optimized code...

For example this bench [https://rawgit.com/thysultan/stylis.js/master/plugins/bytecode/tests/index.html](https://rawgit.com/thysultan/stylis.js/master/plugins/bytecode/tests/index.html), *see the console. It avoids all the points i mentioned above, showing that this bytecode format with a typedArray is faster. ``` 1873 ops, from...

>Can you elaborate on this specific case? The operation is calling a function that returns another function, then `useIt` calls this function. So, you're also measuring the cost of creating...

>Did you see that the first function is self invoking? I missed that :P >Just refreshed a couple of times, the results vary between 2500 and 3500 Same here. >I...

In the bench it's being created on every operation, since the bench runs this operation > 10,000 times it affects the results.

In the real world the data-structure should probably be created just once for the lifetime of the page, the toString implementation on the other hand could get executed more than...

>Sounds like this may have been a browser issue. Possibly still is an issue with requestIdleCallback. >Is it still an issue? Yes, i can still reproduce it.