Thread by @PatrickAlphaC

🏌️‍♀️⛽️Ok EVM gas golfers, got one that is going to test your every gas instinct.

NONE of these functions are optimal... but which one is the most gas efficient and why?

Pick a choice and also leave a comment on why you think so!

❌❌ NO LOOKING ANYTHING UP! ❌❌

a, b, or c?

The answer is C

Each does something a little different gas wise:
- External vs Public functions
- +=
- The location of the function in the file

But the biggest gas factor here is going to be BOTH reading and writing to storage.

The Chainlink core github repo runs with the standard or adding the prefix "s_" to any storage variable, so engineers know it's expensive to use.

In function C, we load the s_playerFunds into memory and read from memory as we loop through the array.

Whereas in a and b, we read right from storage, which is one of the most expensive operations out there!

We also initialize a totalFunds value in memory, instead of constantly writing to s_totalFunds.

a got that part right, but b just wrote to storage for every call.

Long story short:

Reading and writing to storage is expensive! If you have a large array in storage you want to do a lot of reads and writes on, pop it into memory first and go from there!

Here is what that looks like in our function C

playerFunds and totalFunds both get "wiped" after the function finishes, but s_totalFunds and s_playerFunds stays. You can access s_totalFunds & s_playerFunds even after the function finishes, but not playerFunds & totalFunds

And that's the difference between storage vs memory!

Shoot me a like, retweet, and follow if you learned something!

https://twitter.com/PatrickAlphaC/status/1503559512824897540

EDIT: It should be noted, that reading an array in from storage calls SLOAD on every element. But we still get the gas efficiency in C > B because we are calling playerFunds.length instead of s_playerFunds.length