<?xml version="1.0" encoding="UTF-8"?><!-- generator="wordpress/2.1.2" -->
<rss version="2.0" 
	xmlns:content="http://purl.org/rss/1.0/modules/content/">
<channel>
	<title>Comments on: URLQuiz #1: To .WWW or not to .WWW?</title>
	<link>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/</link>
	<description>Advocating User-Centered URL Design</description>
	<pubDate>Fri, 25 Jul 2008 13:57:24 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.1.2</generator>

	<item>
		<title>By: Brad Fults</title>
		<link>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-1460</link>
		<author>Brad Fults</author>
		<pubDate>Mon, 19 Feb 2007 09:36:23 +0000</pubDate>
		<guid>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-1460</guid>
					<description>I vote for option 4.

The "www" subdomain is superfluous and unnecessary baggage on many URLs. Requests to a www subdomain should still work, but simply redirect to the main domain name and serve the appropriate content.

One of the arguments I've heard against deprecation of "www" is that "people won't know it's a web site if they see it in a TV commercial or in a newspaper or hear it on the radio". I say this simply isn't true. People know when they see [something].com or [something].net, etc. that it's an Internet domain name. In fact while visiting Europe I noticed that nearly all advertisements lack a "www" prefix -- it seems the US companies are the slow ones to switch away from this archaic and useless practice.

I see the form of domain names approaching a ubiquity rivaling phone numbers: when consumers see 1-800-XXX-XXXX or 877-XXX-XXXX (in the US at least), they don't have to make sure it's not a Social Security Number or a bank account number; they know it's a phone number.

Similarly (and increasing by the minute), consumers know what a domain name looks like and can eventually navigate to a domain name on the internet without the aid of any silly "www" prefix. This can and probably should be aided by context. With phone numbers it's usually "Call 877-XXX-XXXX", so with domain names it can be "Go to [something].com" or "Visit [something].com".</description>
		<content:encoded><![CDATA[<p>I vote for option 4.</p>
<p>The &#8220;www&#8221; subdomain is superfluous and unnecessary baggage on many URLs. Requests to a www subdomain should still work, but simply redirect to the main domain name and serve the appropriate content.</p>
<p>One of the arguments I&#8217;ve heard against deprecation of &#8220;www&#8221; is that &#8220;people won&#8217;t know it&#8217;s a web site if they see it in a TV commercial or in a newspaper or hear it on the radio&#8221;. I say this simply isn&#8217;t true. People know when they see [something].com or [something].net, etc. that it&#8217;s an Internet domain name. In fact while visiting Europe I noticed that nearly all advertisements lack a &#8220;www&#8221; prefix &#8212; it seems the US companies are the slow ones to switch away from this archaic and useless practice.</p>
<p>I see the form of domain names approaching a ubiquity rivaling phone numbers: when consumers see 1-800-XXX-XXXX or 877-XXX-XXXX (in the US at least), they don&#8217;t have to make sure it&#8217;s not a Social Security Number or a bank account number; they know it&#8217;s a phone number.</p>
<p>Similarly (and increasing by the minute), consumers know what a domain name looks like and can eventually navigate to a domain name on the internet without the aid of any silly &#8220;www&#8221; prefix. This can and probably should be aided by context. With phone numbers it&#8217;s usually &#8220;Call 877-XXX-XXXX&#8221;, so with domain names it can be &#8220;Go to [something].com&#8221; or &#8220;Visit [something].com&#8221;.</p>
]]></content:encoded>
				</item>
	<item>
		<title>By: Edward O'Connor</title>
		<link>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-1538</link>
		<author>Edward O'Connor</author>
		<pubDate>Tue, 20 Feb 2007 17:55:01 +0000</pubDate>
		<guid>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-1538</guid>
					<description>Count me in for #4 as well.</description>
		<content:encoded><![CDATA[<p>Count me in for #4 as well.</p>
]]></content:encoded>
				</item>
	<item>
		<title>By: Devon Young</title>
		<link>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-1540</link>
		<author>Devon Young</author>
		<pubDate>Tue, 20 Feb 2007 18:14:08 +0000</pubDate>
		<guid>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-1540</guid>
					<description>I use a 301 permanant redirect for anything on my site that starts with "www.", since I know a few people who aren't computer literate seem to think that a "www." is a necessary thing. So I doubt a 404 error would make any sense, as it would leave those users confused and they wouldn't even know how to retype the URL so they might get the right place. It's simple, clean, and there's less typing, and no extra "air" filling the bag.</description>
		<content:encoded><![CDATA[<p>I use a 301 permanant redirect for anything on my site that starts with &#8220;www.&#8221;, since I know a few people who aren&#8217;t computer literate seem to think that a &#8220;www.&#8221; is a necessary thing. So I doubt a 404 error would make any sense, as it would leave those users confused and they wouldn&#8217;t even know how to retype the URL so they might get the right place. It&#8217;s simple, clean, and there&#8217;s less typing, and no extra &#8220;air&#8221; filling the bag.</p>
]]></content:encoded>
				</item>
	<item>
		<title>By: Sean Leather</title>
		<link>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-1562</link>
		<author>Sean Leather</author>
		<pubDate>Wed, 21 Feb 2007 02:41:48 +0000</pubDate>
		<guid>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-1562</guid>
					<description>I'm not sure which should be the canonical form. I lean towards no-www (and use this on my sites), because it's less redundant. Unless you have already published a "www." address, I would probably have a 404 and not a redirect, because that leads to "www." URLs in links to your site.</description>
		<content:encoded><![CDATA[<p>I&#8217;m not sure which should be the canonical form. I lean towards no-www (and use this on my sites), because it&#8217;s less redundant. Unless you have already published a &#8220;www.&#8221; address, I would probably have a 404 and not a redirect, because that leads to &#8220;www.&#8221; URLs in links to your site.</p>
]]></content:encoded>
				</item>
	<item>
		<title>By: Igor Clark</title>
		<link>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-1582</link>
		<author>Igor Clark</author>
		<pubDate>Wed, 21 Feb 2007 12:07:46 +0000</pubDate>
		<guid>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-1582</guid>
					<description>I vote for, and implement, #3.

A website is a resource within a domain, and should be addressed as such. There's a sensible organisational reason why DNS has a distinction between A, MX and NS (etc) records, and while it's understandable that some users, notably non-technical ones, don't realise that there is a point to 'www' (or whatever subdomain/hostname is used, which I think is an equally valid but different issue), it's very easy to catch these users via an @ A record and redirect them explicitly via 301 to the canonical 'www' (or other) resource name.

404 would make no sense, as it would just mean a lot of annoyed people not being able to find what they want; 301 would mean people being educated, even if only subliminally, that there is a point to naming conventions and standards. Majority consent is one thing, but the simple fact of a number of people failing to bother to use a standard which is there for a sound structural reason does not mean that the standard should be discarded.</description>
		<content:encoded><![CDATA[<p>I vote for, and implement, #3.</p>
<p>A website is a resource within a domain, and should be addressed as such. There&#8217;s a sensible organisational reason why DNS has a distinction between A, MX and NS (etc) records, and while it&#8217;s understandable that some users, notably non-technical ones, don&#8217;t realise that there is a point to &#8216;www&#8217; (or whatever subdomain/hostname is used, which I think is an equally valid but different issue), it&#8217;s very easy to catch these users via an @ A record and redirect them explicitly via 301 to the canonical &#8216;www&#8217; (or other) resource name.</p>
<p>404 would make no sense, as it would just mean a lot of annoyed people not being able to find what they want; 301 would mean people being educated, even if only subliminally, that there is a point to naming conventions and standards. Majority consent is one thing, but the simple fact of a number of people failing to bother to use a standard which is there for a sound structural reason does not mean that the standard should be discarded.</p>
]]></content:encoded>
				</item>
	<item>
		<title>By: Dan Kubb</title>
		<link>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-1608</link>
		<author>Dan Kubb</author>
		<pubDate>Wed, 21 Feb 2007 21:13:14 +0000</pubDate>
		<guid>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-1608</guid>
					<description>I vote for #4.

I'd take it a step further though and match all subdomains and redirect them to the non-www domain.  That way you catch people who mistype in ww or wwww too.

These days my only usage of subdomains is to set up asset sites (like asset1.example.com, asset2.example.com, etc) to boost the speed of downloading images, javascript and stysleheets.  For more information see:

http://www.ajaxperformance.com/?p=33</description>
		<content:encoded><![CDATA[<p>I vote for #4.</p>
<p>I&#8217;d take it a step further though and match all subdomains and redirect them to the non-www domain.  That way you catch people who mistype in ww or wwww too.</p>
<p>These days my only usage of subdomains is to set up asset sites (like asset1.example.com, asset2.example.com, etc) to boost the speed of downloading images, javascript and stysleheets.  For more information see:</p>
<p><a href="http://www.ajaxperformance.com/?p=33" rel="nofollow">http://www.ajaxperformance.com/?p=33</a></p>
]]></content:encoded>
				</item>
	<item>
		<title>By: vcsjones (ASP.NET forums)</title>
		<link>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-1665</link>
		<author>vcsjones (ASP.NET forums)</author>
		<pubDate>Fri, 23 Feb 2007 03:58:35 +0000</pubDate>
		<guid>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-1665</guid>
					<description>Depends on what the client gives me. Generally I avoid anything implicit, so I know what I am controlling, so #6 is my choicer.</description>
		<content:encoded><![CDATA[<p>Depends on what the client gives me. Generally I avoid anything implicit, so I know what I am controlling, so #6 is my choicer.</p>
]]></content:encoded>
				</item>
	<item>
		<title>By: Haacked</title>
		<link>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-2050</link>
		<author>Haacked</author>
		<pubDate>Thu, 01 Mar 2007 07:21:01 +0000</pubDate>
		<guid>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-2050</guid>
					<description>Well I think you know where I stand on this, I'm going to have to go for #4. Something I have yet to implement for my own blog.</description>
		<content:encoded><![CDATA[<p>Well I think you know where I stand on this, I&#8217;m going to have to go for #4. Something I have yet to implement for my own blog.</p>
]]></content:encoded>
				</item>
	<item>
		<title>By: simone</title>
		<link>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-2059</link>
		<author>simone</author>
		<pubDate>Thu, 01 Mar 2007 11:02:41 +0000</pubDate>
		<guid>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-2059</guid>
					<description>My vote is for #3.

I like the www since I started working on the web in 1995, so I like it for historical reason...
but the root must also be accepted, but having both raises many problem with cookies and stats as they are 2 different domains for the same site</description>
		<content:encoded><![CDATA[<p>My vote is for #3.</p>
<p>I like the www since I started working on the web in 1995, so I like it for historical reason&#8230;<br />
but the root must also be accepted, but having both raises many problem with cookies and stats as they are 2 different domains for the same site</p>
]]></content:encoded>
				</item>
	<item>
		<title>By: Carl Camera</title>
		<link>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-2072</link>
		<author>Carl Camera</author>
		<pubDate>Thu, 01 Mar 2007 14:54:05 +0000</pubDate>
		<guid>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-2072</guid>
					<description>Another vote for 4 from me, but I have not implemented.  In addition, the Google Webmaster Tools interface allows me to specify my preferred canonical form (Domain Preference tab) for my websites -- which I have done.</description>
		<content:encoded><![CDATA[<p>Another vote for 4 from me, but I have not implemented.  In addition, the Google Webmaster Tools interface allows me to specify my preferred canonical form (Domain Preference tab) for my websites &#8212; which I have done.</p>
]]></content:encoded>
				</item>
	<item>
		<title>By: Alan PInstein</title>
		<link>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-4179</link>
		<author>Alan PInstein</author>
		<pubDate>Sat, 31 Mar 2007 01:46:36 +0000</pubDate>
		<guid>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-4179</guid>
					<description>Obviously they both need to work. I cringe every time I go to a company's website WITHOUT the www and see an (invariably IIS) 404 for the domain. 

Long-term I favor just domain.com as it's less to type and "teaches" people that www is not meaningful. People need to realize this. 

My company offers a CMS for the real estate industry. We have generally unsopshisticated computer users as customers. They are quite confused when I try to get them to go with a web site without the www in front. For instance, if I try to get them to go to a subdomain, as in "Go to subdomain.showcasere.com in your browser", they always instantly respond with "www. what?". And then I say "No www", and then they end up typing in "subdomain@showcasere.com". Seemingly if it doesn't start with "www" some people still just don't recognize it as a web site. It's very frustrating.

Our system produces PDF flyers for houses, and we give each property its own address, such as "123MainStreet.ViewHouseForSale.com". I left the www off for brevity. Unfortunately, customers were confused as to "what is that thing on the flyer". 

So that should let you know at least anecdotally that people aren't ready for no WWW.

And I'd certainly recommend that any site that accepts subdomains *also* accept www. in front of *anything* and work correctly.</description>
		<content:encoded><![CDATA[<p>Obviously they both need to work. I cringe every time I go to a company&#8217;s website WITHOUT the www and see an (invariably IIS) 404 for the domain. </p>
<p>Long-term I favor just domain.com as it&#8217;s less to type and &#8220;teaches&#8221; people that www is not meaningful. People need to realize this. </p>
<p>My company offers a CMS for the real estate industry. We have generally unsopshisticated computer users as customers. They are quite confused when I try to get them to go with a web site without the www in front. For instance, if I try to get them to go to a subdomain, as in &#8220;Go to subdomain.showcasere.com in your browser&#8221;, they always instantly respond with &#8220;www. what?&#8221;. And then I say &#8220;No www&#8221;, and then they end up typing in &#8220;subdomain@showcasere.com&#8221;. Seemingly if it doesn&#8217;t start with &#8220;www&#8221; some people still just don&#8217;t recognize it as a web site. It&#8217;s very frustrating.</p>
<p>Our system produces PDF flyers for houses, and we give each property its own address, such as &#8220;123MainStreet.ViewHouseForSale.com&#8221;. I left the www off for brevity. Unfortunately, customers were confused as to &#8220;what is that thing on the flyer&#8221;. </p>
<p>So that should let you know at least anecdotally that people aren&#8217;t ready for no <a href="http://WWW." rel="nofollow">WWW.</a></p>
<p>And I&#8217;d certainly recommend that any site that accepts subdomains *also* accept <a href="http://www." rel="nofollow">www.</a> in front of *anything* and work correctly.</p>
]]></content:encoded>
				</item>
	<item>
		<title>By: Mark Andrews</title>
		<link>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-4459</link>
		<author>Mark Andrews</author>
		<pubDate>Mon, 02 Apr 2007 06:06:42 +0000</pubDate>
		<guid>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-4459</guid>
					<description>There really is nothing new in the world.  Just old ideas revamped. www will go the same way that host labels (and sub domains) have gone from email addresses.

However for this to achieve ubiquity we need to address some of the day-to-day management issues required to do this.

Today that solution is to add address (A/AAAA) records at the apex of the zone.  This however is a non-optimal solution especially if the site is being hosted by a second party.

* There is a slight risk that old  MTA's will attempt to deliver email to the hosting machines.
* It's hard for the hosting sites to change the addresses of the servers as it requires a large co-ordinated effort.
* The usual instructions for www to add a CNAME don't work for names which are the apex of zones. CNAME are illegal there.

There is however a potential solution which is to use SRV records (RFC 2782).  These provide a level of indirection (which CNAME provide for www).  They also don't run the risk of misdirected email. They can be used regardless of where the zone cuts occur in the DNS hierarchy.

http://www.watersprings.org/pub/id/draft-andrews-http-srv-01.txt</description>
		<content:encoded><![CDATA[<p>There really is nothing new in the world.  Just old ideas revamped. www will go the same way that host labels (and sub domains) have gone from email addresses.</p>
<p>However for this to achieve ubiquity we need to address some of the day-to-day management issues required to do this.</p>
<p>Today that solution is to add address (A/AAAA) records at the apex of the zone.  This however is a non-optimal solution especially if the site is being hosted by a second party.</p>
<p>* There is a slight risk that old  MTA&#8217;s will attempt to deliver email to the hosting machines.<br />
* It&#8217;s hard for the hosting sites to change the addresses of the servers as it requires a large co-ordinated effort.<br />
* The usual instructions for www to add a CNAME don&#8217;t work for names which are the apex of zones. CNAME are illegal there.</p>
<p>There is however a potential solution which is to use SRV records (RFC 2782).  These provide a level of indirection (which CNAME provide for www).  They also don&#8217;t run the risk of misdirected email. They can be used regardless of where the zone cuts occur in the DNS hierarchy.</p>
<p><a href="http://www.watersprings.org/pub/id/draft-andrews-http-srv-01.txt" rel="nofollow">http://www.watersprings.org/pub/id/draft-andrews-http-srv-01.txt</a></p>
]]></content:encoded>
				</item>
	<item>
		<title>By: biophonc</title>
		<link>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-8945</link>
		<author>biophonc</author>
		<pubDate>Tue, 08 May 2007 15:51:29 +0000</pubDate>
		<guid>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-8945</guid>
					<description>#4
imho it's the best solution for me atm. 

I do not like the www thing at all, but a lot of big sites use the www thing in online and offline marketing and also have only the www sub domain active/in use and or make a redirect to www. However,... for most people the www prefix stands for a link to a website. Without the www prefix they think the website does not work. www.*.tld = entity ? They always look irritated when I tell them it's absolutely not necessary but they do not understand(or don't want) and type again www.*.tld

The best thing would be (obviously) to use #2 to reeducate the (l)user :)</description>
		<content:encoded><![CDATA[<p>#4<br />
imho it&#8217;s the best solution for me atm. </p>
<p>I do not like the www thing at all, but a lot of big sites use the www thing in online and offline marketing and also have only the www sub domain active/in use and or make a redirect to <a href="http://www." rel="nofollow">www.</a> However,&#8230; for most people the www prefix stands for a link to a website. Without the www prefix they think the website does not work. <a href="http://www." rel="nofollow">www.</a>*.tld = entity ? They always look irritated when I tell them it&#8217;s absolutely not necessary but they do not understand(or don&#8217;t want) and type again <a href="http://www." rel="nofollow">www.</a>*.tld</p>
<p>The best thing would be (obviously) to use #2 to reeducate the (l)user :)</p>
]]></content:encoded>
				</item>
	<item>
		<title>By: Johannes Rössel</title>
		<link>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-83372</link>
		<author>Johannes Rössel</author>
		<pubDate>Thu, 01 May 2008 12:03:37 +0000</pubDate>
		<guid>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-83372</guid>
					<description>As for my site the default configuration was that www and non-www pointed to the same content. After a few days the site was on I noticed that some search engines indexed the www variant, some the non-www variant and iirc one even treated both differently and yielded results with and without the www. I decided pretty early (so to not break many links, if there ever were any ...) to get rid of the www form and installed a redirect from www. to the non-www form (unfortunately only for the root, so no deep links were considered. However I never got around thinking about how I might do that :).

I also noticed that I almost never type the www anymore, anyway. Except for a few sites I know where I get nothing without the www and they constantly annoy me. It's also not nice to the user to issue 404s in that case, imho.</description>
		<content:encoded><![CDATA[<p>As for my site the default configuration was that www and non-www pointed to the same content. After a few days the site was on I noticed that some search engines indexed the www variant, some the non-www variant and iirc one even treated both differently and yielded results with and without the <a href="http://www." rel="nofollow">www.</a> I decided pretty early (so to not break many links, if there ever were any &#8230;) to get rid of the www form and installed a redirect from <a href="http://www." rel="nofollow">www.</a> to the non-www form (unfortunately only for the root, so no deep links were considered. However I never got around thinking about how I might do that :).</p>
<p>I also noticed that I almost never type the www anymore, anyway. Except for a few sites I know where I get nothing without the www and they constantly annoy me. It&#8217;s also not nice to the user to issue 404s in that case, imho.</p>
]]></content:encoded>
				</item>
	<item>
		<title>By: The Great Dub-Dub-Dub Debate &#124; Developer Home</title>
		<link>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-89955</link>
		<author>The Great Dub-Dub-Dub Debate &#124; Developer Home</author>
		<pubDate>Thu, 22 May 2008 22:14:30 +0000</pubDate>
		<guid>http://blog.welldesignedurls.org/2007/02/19/urlquiz-1-www-or-non-www/#comment-89955</guid>
					<description>[...] that, it&#8217;s largely a matter of taste, though you could make a case that user-centered URL design should rule the day. If we&#8217;re dropping the www prefix, why stop there? Why not drop the [...]</description>
		<content:encoded><![CDATA[<p>[&#8230;] that, it&#8217;s largely a matter of taste, though you could make a case that user-centered URL design should rule the day. If we&#8217;re dropping the www prefix, why stop there? Why not drop the [&#8230;]</p>
]]></content:encoded>
				</item>
</channel>
</rss>
