Precision Computing

Could you explain why "This is a bloated and slow way to do a for loop" because the Hankatsu person seems to think that there is no drawback (or fast).

I meant to add the URL for reference:

http://d.hatena.ne.jp/hankatsu/20050729/1122659503

Sorry, Clifford -- I still can't read what Hankatsu wrote. He or she quoted this post, then expanded it a little bit to make a cool download script.

The reason that this is bloated and slow is that the "1..5" expression actually creates an array of 5 numbers, then passes it down the pipeline. Next, the foreach does one thing for each number in the array.

You can see the efficiency differences with time-expression:

MSH:2 C:\Temp >time-expression { 1..10000 | foreach {} }
TotalMilliseconds : 1078.125

MSH:3 C:\Temp >time-expression { foreach ($x in 1..10000) {} }
TotalMilliseconds : 15.625

One important thing to remember about performance optimizations, though, is that they can always change under you. It is possible for us to make the "bloated and slow" version nearly as fast as the second version, but it probably won't reduce its memory consumption.

Also, with all development -- it's best to program to your intention, not to side-effects. On the command-line, take all of the typing short cuts you want. In a script, though, non-intuitive shortcuts often make your script difficult to read and error prone.

Lee