From edi@agharta.de Thu Jul 27 17:59:46 2006
From: Edi Weitz <edi@agharta.de>
To: cl-who-announce@common-lisp.net
Subject: [cl-who-announce] New release 0.6.1 (Was: CL-WHO:STR behavior and
	package nicknames.)
Date: Thu, 27 Jul 2006 23:51:41 +0200
Message-ID: <u7j1yemxe.fsf@agharta.de>
In-Reply-To: <58f839b70607271337y2e0dec65s975d7d2ab020795d@mail.gmail.com>
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="===============3904192438046883590=="

--===============3904192438046883590==
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: 7bit

Hi Erik!

On Thu, 27 Jul 2006 16:37:02 -0400, "Erik Enge" <erik.enge(a)gmail.com> wrote:

> In the documentation I find the following example:
>
> (:table :border (+ 1 2)) => (write-string "<table border='" s)
>                                              (princ (+ 1 2) s)
>                                                (write-string "' />" s)
>
> However, in my implementation I see the following:
>
> CL-USER> (cl-who:with-html-output-to-string (*standard-output*)
>                     (:table :border (+ 1 2)))
> "<table border='(+ 1 2)' />"
>
> This makes sense because CONSTANTP returns T for that form in SBCL
> and NIL in CMUCL.

Right, I didn't think of that.  What I actually /meant/ in this case
was that this form should be evaluated and I think because it is
CONSTANTP this is one of the few occasions where usage of EVAL is OK.
Anyway, I've uploaded a new version which does exactly that.

> In which case I think I'm supposed to use the STR operator:

No, STR, ESC, and FMT aren't really meaningful in attribute positions.
They are supposed to occur in the body of a tag.

> Separately, would you consider adding the nickname "who" (or
> whatever you like, just shorter than "cl-who") to the package?

Done in 0.6.1.

Thanks for the report,
Edi.

--===============3904192438046883590==--