W3C home > Mailing lists > Public > www-html@w3.org > February 2004

Target attrib in HTML 4.01 Strict

From: Justin Wood <Callek@juno.com>
Date: Sat, 07 Feb 2004 13:00:26 -0500
Message-ID: <402527BA.8020104@juno.com>
To: www-html@w3.org
Cc: Videolego@aol.com

(warning a bit long)

I firstly want to say, I did take the time to scim through the archives 
of this list to ensure I don't frustrate all of you with a very repeated 

Though this has been brought up before, I will cite old(er) messages and 
provide counter arguments, from both sides (I hope).

It is in my opinion that the target attribute should be in HTML 4.01 
strict with the following guideline, only the built-in uses for target 
should be recognized, (_blank, _top, etc.)

It is my understanding there is currently no CSS1, CSS2, or proposed 
CSS3 property to achieve this.  If I am wrong would someone please 
correct me, and I will retract my statement(s) here.

each source will be citied with a certain number of *'s and refferenced 
at the bottom of this mail.

I feel Joe Kaczmarek (*)  had it correct, target does not neccessarily 
mean "Frames", there could be possibilities for non-visual UA's to open 
a _blank window/interface, to their discretion...

 > It's my understanding that targetting, opening windows and so on is
 > putatively the realm of scripting/DOM, not hard-coded HTML semantics.
 > Thus, with proper UA support for HTML's OBJECT element, DOM, and
 > CSS-2, frames will be truly obsolete in every respect. We're not
 > there yet, of course.
~Todd Fahrner (**)

The realm of Scripting/DOM to me, is more a way to handle user 
inter-action and prefferences in a certain page.  Proper UA support for 
the methods 'target' (attrib) currently supports by means of things like 
_blank, and _top do not mean anything if the WG for CSS and/or HTML do 
not provide alternate means. There are some TABLE properties that were 
not deprecated since CSS 1 does not have the methods to support all of 
it, as is said in the spec.

 >< Joe Kaczmarek's full mail (***)

That is one of the big issues for the 'target' attrib, if a user creates 
a FRAMESET for themselves, suppose their UA supports frames but not 
bookmarks (hypothetical) so they manually create a framset doc with the 
homepage as the first line, a form's input line and a bit JS to 
navigate, and a line or two for their favorites, who are we to have a 
script run that says "if (top!=self) ..."  that would be very 
frustrating to the end user, I for one, never want a page I visit to 
AUTO-break out of a frame, I want the option. 

which brings me to my next cite:

 > Why not do something like this?
 > <a href="foobar.html" onclick="open_new_window('foobar.html')">Sample
 > web site</a>
~Jim Correia (****)

The reason why not, is partly like I said above, and what about those 
PDA's for instance, (or tablet PC's for that matter) which DO support 
multiple windows, but however since there is no mouse, DO NOT support 
onclick, since if the UA does not have access to a device, such as a 
mouse, they can still be conformant without supporting it's methods.

 >< see two sources (*|*) (*|**)

I agree with most of what was said, but shouldnt HTML default to, "allow 
unless CSS has an alternative), I can see a bit more lee-way in this 
idea with XHTML and XML since they  allow alternate methods around the 

 >< David J Woolley [DJW] (*|***)

Sure it can be abused, but what about all the legitimate reasons for it.

I for one, have had a habit of leaving a link on my index.html 
(entrance) page to allow users to break out of another websites frame, 
and on any external link I have had a habit of using target _blank with 
a note that it opens in a new window.

When I visit a website, that is my expected behavior, I get annoyed when 
the href of a link is Javascript, I get annoyed when I expect to still 
be in one site, and click an external link and am no longer in that 
site, but in the same window, makes for confusing practice.

Also as a website designer, when you have links elsewhere, you _want_ 
your users/viewers to have (quick) access to your page after they click 
a pop-up link, not everyone even knows there is a 'back' button.  When 
getting deep into another wbsites architechture the back-button can even 
become annoying to some.  All I ask is for functionality to be legal in 
Strict at least until it is presented in (a public) CSS format.

Comments/Questions welcome, I think I covered all original points (which 
mostly have been repeated on this list) as well as provide new points of 

~Justin Wood
(*) -  Joe Kaczmarek -- Fri, 05 May 2000 17:17:46-0400
 -- Re: HTML 4.01 Strict DTD missing an important attribute
 -- http://lists.w3.org/Archives/Public/www-html/2000May/0039.html
(**) - Todd Fahrner -- Fri, 5 May 2000 14:36:19-0700
 -- Re: HTML 4.01 Strict DTD missing an important attribute
 -- http://lists.w3.org/Archives/Public/www-html/2000May/0040.html
(***) - Joe Kaczmarek -- Mon, 28 Aug 2000 11:08:46-0400 (EDT)
 -- Target attribute in STRICT DTD
 -- http://lists.w3.org/Archives/Public/www-html/2000Aug/0116.html
(****) - Jim Correia -- Fri, 1 Sep 2000 21:18:32-0400
 -- RE: the 'target' attribute
 -- http://lists.w3.org/Archives/Public/www-html/2000Sep/0004.html
(*|*) - Timothy J. Luoma -- Sun, 3 Mar 2002 00:24:20-0500 (Eastern 
Standard Time)
 -- Re: Target, or opening a new instance of a UA with a link
 -- http://lists.w3.org/Archives/Public/www-html/2002Mar/0007.html
(*|**) - Thomas Hurst -- Sun, 3 Mar 2002 17:08:52+0000
 -- Re: Target, or opening a new instance of a UA with a link
 -- http://lists.w3.org/Archives/Public/www-html/2002Mar/0009.html
(*|***) - David J Woolley -- Wed, 30 Aug 2000 19:37:22+0100
 -- RE: the 'target' attribute
 -- http://lists.w3.org/Archives/Public/www-html/2000Aug/0120.html
Received on Saturday, 7 February 2004 13:12:12 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:06:07 UTC