Re: bad-idea-of-the-day: Inline data as URL scheme?

Terry Allen (
Fri, 1 Sep 1995 09:00:57 -0700

Re SGML tools, we found one that respected the limits in the SGML
decl awhile ago. OLIAS? I agree the limits (like capacities) aren't
useful, but that's something to tell to ANSI, where I understand
SGML revision efforts are rooted.

--- Forwarded mail from Gavin Nicol <>

Resent-Date: Fri, 1 Sep 1995 11:49:58 -0400
Date: Fri, 1 Sep 1995 11:53:22 -0400
From: Gavin Nicol <>
In-Reply-To: <> (
Subject: Re: bad-idea-of-the-day: Inline data as URL scheme?
X-Mailing-List: <> archive/latest/1714
Precedence: list

Dan says:
>Er... this limit is (a) not a good idea, (b) not enforced except
>in things like SGMLS. I think it should go away (or become 999999,
>at least).

I agree with (a) most emphatically, but (b) is false. I remember
testing the limits on many browsers about 7 months ago, and found
that all had limits (some smaller than 1024), and some even crashed
when the buffers overflowed. Does anyone have new information?

--- End of forwarded mail from Gavin Nicol <>


Terry Allen  (   O'Reilly & Associates, Inc.
Editor, Digital Media Group    101 Morris St.
			       Sebastopol, Calif., 95472

A Davenport Group sponsor. For information on the Davenport Group see or Current HTML 2.0 spec: