Question

I’m currently in a low level C class, but I program Javascript as my job.

Here are a couple C optimizations/Problem areas my book has gone over so far (Just started so I don’t have much):

• Procedure overhead (restoring registers, creating the stack, etc.)
• Sometimes creating temporary variables instead of reading/writing to memory in loops.

Additionally how much time can be spent ‘optimizing’ Javascript? C is compiled down to machine code and much more time can be spent on compiling an executable. How much time is spent in a typical JS compiler optimizing?

Javascript is a higher level language than C and because of that I don’t actually ‘know’ what is going on underneath. What optimizations should I worry about? Will what I am learning now in C also apply in Javascript?

Answer 1

Javascript and C have such radically different ways of expressing things that I doubt one could even compare the optimization techniques.

For this post, I’ll just look at the optimizations the programmer can make:

C:

• Choosing the most appropriate data type. (int/long/float/double)
• Bit fields and aligning on word boundaries.
• Use of malloc and choosing when to free memory and when to reuse it.
• Intermediate calcuations

Javascript:

• Avoid creating pointless objects when needed. Object re-use
• Intermediate calculations
• Avoiding refreshing the DOM too often
• Set big unused references to null

Just as a quick example. The two really are worlds apart, so I’d have to simplify optimization techniques down to:

• Reuse what you can
• Don’t make what you don’t have to.
• If you do something computationally expensive and you need it again — save it.
• Choose an appropriate algorithm.

The last item on the list is, as far as I am concerned, the most import bit of optimization that any coder can learn how to do.

• Don’t use the BogoSort if you have a better sort to use. (Hint, you do.)
• See if your standard library has a fast way of doing what you want to do. Don’t re-invent the wheel until you proved your old wheel doesn’t work.
• Don’t worry too much about doing micro-optimization until profiling shows you it is worth the effort.

Disclaimer: Yes, I know the old adage of ‘Premature optimization is the root of all evil.’ Choosing a smart, or at the very least, not terribly dumb way to solve a problem is not premature..