Does wmic really output UTF-8?

It appears it does (some form of unicode anyway), if you redirect its output to a file (or via a pipe):

c:\test>wmic.exe process > junk

c:\test>u:head -c 200 junk
 ■C a p t i o n                             C o m m a n d L i n 
+e
[download]

That first spludge(*) is the LE BOM 0xfffe. (It was a spludge when I c&p'd it!)

Probably utf-16le I think, but whatever it is, using -CS does seem to cause perl to work out what it is getting and treat it appropriately:

c:\test>wmic.exe process | perl -CS -pe1 > junk

c:\test>u:head -c 200 junk
Caption              CommandLine
[download]

Whether by accident or design, it is useful.

I didn't suggest
wmic.exe process > junk
I suggested you drop the -CS, so you need to compare
wmic.exe process | perl -CS -pe1 > junk
to
wmic.exe process | perl -pe1 > junk

It's simple,

• If wmic outputs "C a p t i o n", then it's cleary using a 16-bit encoding and -CS is guaranteed to hurt.
• If wmic doesn't output "C a p t i o n", then you've lost your justification for using -CS.

Now I'm confused. a) how would it hurt? b) Why does it seem not to hurt in my use now?

---

#my sig used to say 'I humbly seek wisdom. '. Now it says:
use strict;
use warnings;
I humbly seek wisdom.