From - Tue Jun 29 16:22:50 1999
Received: from newton.math.grin.edu ([132.161.33.70]) by EXCHANGESERVER with SMTP (Microsoft Exchange Internet Mail Service Version 5.5.2448.0)
	id NK81YCA6; Tue, 29 Jun 1999 09:45:38 -0500
Received: from CS.Princeton.EDU (CS.Princeton.EDU [128.112.136.10])
	by newton.math.grin.edu (8.8.8/8.8.8) with ESMTP id JAA11297
	for <rebelsky@math.grin.edu>; Tue, 29 Jun 1999 09:47:03 -0500 (CDT)
Received: from ux04 (appel@ux04.CS.Princeton.EDU [128.112.169.24])
	by CS.Princeton.EDU (8.8.8/8.8.8) with ESMTP id KAA15576
	for <rebelsky@math.grin.edu>; Tue, 29 Jun 1999 10:44:23 -0400 (EDT)
Received: from localhost by ux04 (8.8.8+Sun) id KAA15607; Tue, 29 Jun 1999 10:44:23 -0400 (EDT)
Message-Id: <199906291444.KAA15607@ux04>
To: rebelsky@math.grin.edu
Subject: Re: MCI Bug Report 
In-reply-to: Your message of "Mon, 12 Oct 1998 12:38:32 EDT."
             <199810121638.MAA14347@mnemonic> 
Date: Tue, 29 Jun 1999 10:44:23 -0400
From: Andrew Appel <appel@CS.Princeton.EDU>

Dear Prof. Rebelsky:

Thanks for this bug report.

		-- Andrew Appel

> 
> The sample let expression on p. 99 appears to be semantically incorrect.
> That expression reads
> 
> let var a := 5
>     function f() : int = g(a)
>     function g(i: int) = f()
>   in f()
> end
> 
> However, according to the definition of Tiger, g is a procedure and therefore
> does not return a result.  Hence, the body of f cannot consist of only
> a call to g.
> 
> The same invalid expression also appears in the Java version of the text.