Is faster using abbreviated parameter names?

ruby-talk
1

Hi, a simple question:

a)
  def hello(string)
    # stuff with 'string'
  end

b)
  def hello(s)
    # stuff with 's'
  end

Will be b) faster since the parameter name contains less letters?

···

--
Iñaki Baz Castillo

2

I have no idea. I guess not. But you can easily verify for yourself. Just write up a little program using Benchmark and you'll soon know.

Kind regards

  robert

···

On 02.03.2009 21:48, Iñaki Baz Castillo wrote:

Hi, a simple question:

a)
  def hello(string)
    # stuff with 'string'
  end

b)
  def hello(s)
    # stuff with 's'
  end

Will be b) faster since the parameter name contains less letters?

3

Yes, doing a benchmark the result is more or less the same (any other factor
seems to be more important), but what I want to know is what should be the
response based on how Ruby works. Under my understanding Ruby needs to parse
during runtime the variable name so a longer variable name would require more
time, am I wrong?

Thanks a lot.

···

El Lunes, 2 de Marzo de 2009, Robert Klemme escribió:

On 02.03.2009 21:48, Iñaki Baz Castillo wrote:
> Hi, a simple question:
>
> a)
> def hello(string)
> # stuff with 'string'
> end
>
> b)
> def hello(s)
> # stuff with 's'
> end
>
>
> Will be b) faster since the parameter name contains less letters?

I have no idea. I guess not. But you can easily verify for yourself.
Just write up a little program using Benchmark and you'll soon know.

--
Iñaki Baz Castillo

4

And how much faster would it have to be to make it worth it to use a meaningless name versus a meaningful one?

Smells like premature optimization to me...

Cheers,

dwh

Robert Klemme wrote:

···

On 02.03.2009 21:48, Iñaki Baz Castillo wrote:

Hi, a simple question:

a)
  def hello(string)
    # stuff with 'string'
  end

b)
  def hello(s)
    # stuff with 's'
  end

Will be b) faster since the parameter name contains less letters?

5

Iñaki Baz Castillo wrote:

Yes, doing a benchmark the result is more or less the same (any other
factor
seems to be more important), but what I want to know is what should be
the
response based on how Ruby works. Under my understanding Ruby needs to
parse
during runtime the variable name so a longer variable name would require
more
time, am I wrong?

I believe you're wrong.

Symbols are resolved into references to the symbol table at parse time,
so when running, :s and :ssssssssssssssss are just two different
pointers into the same symbol table. As for local variables, they are
just offsets into the stack frame.

So it might take a microscopically small amount of extra time for your
program to start up, reading a few extra bytes of source code, but once
it's running, each iteration should take the same time.

···

--
Posted via http://www.ruby-forum.com/.

6

Sure, it was just curiosity :slight_smile:
I will not code in Ruby to get a unreadable code like in other languages :slight_smile:

···

El Lunes, 2 de Marzo de 2009, Denis Haskin escribió:

And how much faster would it have to be to make it worth it to use a
meaningless name versus a meaningful one?

Smells like premature optimization to me...

--
Iñaki Baz Castillo

7

Thanks, that makes sense :slight_smile:

···

El Lunes, 2 de Marzo de 2009, Brian Candler escribió:

Iñaki Baz Castillo wrote:
> Yes, doing a benchmark the result is more or less the same (any other
> factor
> seems to be more important), but what I want to know is what should be
> the
> response based on how Ruby works. Under my understanding Ruby needs to
> parse
> during runtime the variable name so a longer variable name would require
> more
> time, am I wrong?

I believe you're wrong.

Symbols are resolved into references to the symbol table at parse time,
so when running, :s and :ssssssssssssssss are just two different
pointers into the same symbol table. As for local variables, they are
just offsets into the stack frame.

So it might take a microscopically small amount of extra time for your
program to start up, reading a few extra bytes of source code, but once
it's running, each iteration should take the same time.

--
Iñaki Baz Castillo

8

Good to hear.

It's just remarkable how lots of people start at the end and work their way backwards... :wink:

dwh

Iñaki Baz Castillo wrote:

···

El Lunes, 2 de Marzo de 2009, Denis Haskin escribió:
  

And how much faster would it have to be to make it worth it to use a
meaningless name versus a meaningful one?

Smells like premature optimization to me...
    
Sure, it was just curiosity :slight_smile:
I will not code in Ruby to get a unreadable code like in other languages :slight_smile: