What good is ``?

What good is ``?

[Mark-Jason Dominus]

On a related note: Given Byron's policy of not adding features unless
the feature implements genuinely new functionality, what's the rationale
behind ``?

Why not replace
	`` (foo) {bar}

with

	`{ ifs=(foo) echo `{bar}}

which is the obvious way to do the same thing without a crufty
piece of syntax?

People who like `` anyway can make a shell function.

  The night is pleasing to us because, like memory, it erases idle details.
Mark-Jason Dominus 	  			    mjd@central.cis.upenn.edu 

Re: What good is ``?

[Paul Haahr]

i guess i should respond to this, since i pressured byron into adding ``
many moons ago.  my argument is you need either ` and ifs, or `` and
`` is simpler.  my best of all possible worlds solution gets rid of
ifs, because i don't like state variables.  with ifs, how do you use
two different backquote expressions in the same command?  (answer:
you can't.)

incidentally, none of these
	x=`{ ifs=: echo -n a:b }
	x=` ifs=: { echo -n a:b }
	ifs=: { x=`{ echo -n a:b } }
is the same as
	x=`` : { echo -n a:b }
although this is:
	ifs=: { x=`{ echo -n a:b } ; }

moreover, it is impossible to define a shell function to do `` because
brace enclosed groups are not first class objects in rc.  (that is my
most serious gripe against the language.)

`` may very well be a mistake.  certainly it and ifs are redundant.  i
believe byron's implementation translates
	`foo
to
	`` $ifs foo
internally, and vice-versa for whatis, so the cost seems minimal to me.
it depends on which one you treat as extra.  (that ifs is in td's paper
probably makes that a closed issue, unfortunately for my case.)

so what do other people say?  is `` too much of a blemish?

paul

[ it is surprising to me that i find myself being the main proponent of
creeping featurism.  i'm usually a vociferous opponent.  i guess that's
what happens when you become a user instead of a designer. ]