Question Ascii in triggers

I'm working on a fishing trigger and I decided to make some cute lil waves to indicate the passing time as you fish. As always, I tested this trigger in speech form first.

1) Name : test trigger
2) Intended for : Rooms
3) Trigger types: Speech
4) Numeric Arg : 100
5) Arguments : test
6) Commands:
%send% %actor% You cast your line into the waters.
wait 1 second
%echo% @c ~~ ~~ ~ ~ @n
....
/t
/fi

This shows as:
You cast your line into the waters.
~~ ~~ ~ ~ (except in cyan)

Then I switch to an object trigger:
1) Name : Fishing
2) Intended for : Objects
3) Trigger types: Command
4) Numeric Arg : 4
5) Arguments : fish
6) Commands:
%send% %actor% You cast your line into the waters.
wait 1 second
%echo% @c ~~ ~~ ~ ~ @n
....
/t
/fi

This shows as:
You cast your line into the waters.
someonesomeonesomeonesomeonesomeonesomeone

I can get around this by inputting it this way:
%echo% @c \~ \~\~ \~ \~ @n

But I'm just wondering...um, why?

Internally, %echo% do some substitutions.
github.com/tbamud/tbamud/blob/master/src/dg_comm.c#L47 This will replace in the name or personal pronoun, depending on the token and the trigger type.
I'm unsure if this is documented anywhere...

Actually, I'm not so surprised at someonesomeonesone, I'm surprised that the speech trigger doesn't do it. The room speech does exactly what *I* told it to but it shouldn't have, should it? The object command does as the *code* tells it and substitutes someone unless an escape...um, code? sequence? So what is in that room speech trigger that doesn't need the escape? Is it this part?

thomas wrote:

Code:

else strcat(sb,PERS((char_data *)otokens[i], ch)); break; ...

In my weird non-coder way, I'm seeing: if there's a token, substitute someone and if there isn't, substitute you. And maybe: if there isn't a token and there isn't not a token don't do anything? Um, what's a token, please?