Re: NRE committed: PLEASE TEST
- From: john roll <john@xxxxxxxxxx>
- Date: Tue, 19 Aug 2008 06:44:31 -0700 (PDT)
Miguel, Donal,
I'm very appreciative of these new features going in to the core, and
of your efforts. Tcl has such a clean look (to me) that I want to put
my input on the table early. I know that it is quite difficult to
hide some things due to Tcl's lack of syntax. Where other languages
have special constructs or know something special about a keyword, we
just have commands.
I see in the above example that you've moved the initial yield into
the body of the coroutine call to clear it out of the iota routine
proper, but I'm asking the more fundamental question, why have an
initial execution of the script passed to the coroutine command at
all? Why not just build the coroutine command immediately? Will this
feature be useful elsewhere?
When I initially suggested the possible usage :
while { [set x [iota 2 20]] } { puts $x }
I was going for the feeling of Icon, but I realized after I posted
that Tcl doesn't have the special value that Icon calls failure. The
"while { 1 } { set x [$iota] }" expression is almost as good but it is
less suggestive of what is controlling the loop termination.
Thanks,
John
.
- Follow-Ups:
- Re: NRE committed: PLEASE TEST
- From: Fredderic
- Re: NRE committed: PLEASE TEST
- References:
- Re: NRE committed: PLEASE TEST
- From: john roll
- Re: NRE committed: PLEASE TEST
- From: miguel
- Re: NRE committed: PLEASE TEST
- From: john roll
- Re: NRE committed: PLEASE TEST
- From: Donal K. Fellows
- Re: NRE committed: PLEASE TEST
- Prev by Date: Re: Running BASH script from TCL
- Next by Date: Re: Tcl/Tk 8.5.4 RELEASED and long-known regression is still there :(
- Previous by thread: Re: NRE committed: PLEASE TEST
- Next by thread: Re: NRE committed: PLEASE TEST
- Index(es):
Relevant Pages
|