W3C home > Mailing lists > Public > public-egov-ig@w3.org > November 2010

RE: Does the user care about URLs?

From: Gannon Dick <gannon_dick@yahoo.com>
Date: Wed, 10 Nov 2010 11:48:42 -0800 (PST)
Message-ID: <63152.19690.qm@web112601.mail.gq1.yahoo.com>
To: rachel.flagg@gsa.gov, Hugh Barnes <Hugh.Barnes@nehta.gov.au>
Cc: chris@e-beer.net.au, W3C e-Gov IG <public-egov-ig@w3.org>
Hugh, you raise some interesting points.

There is a transportation model buried in URL's and URI's that might provide some guidance.  It may not have been the overt intentions of the IETF, but a system that "works" has some features.

If I were Maxwell's Demon I would take note of the following:

1) The web has "Origin-less" addressing, meaning that a URL/URI is a right directed graph with the protocol to the far left and the destination to the right.  Only the mode of transportation you use to reach the data matters, not the path or length of the path.  Maxwell's Demon cannot deliver data because it (only) has an Origin from which to serve data.

2) IRL, an International Air Waybill has 12 characters, Sea Waybills have more characters, but the de facto standard is 12.  This means (to the Demon) that any incoming deliveries with less than 19 characters (http:/)(/)(12CHAR) are of local origin.  Local to what ?  That is indeterminate.

3) If the Demon is selling the data, the only applicable INCO terms would be FAS, FOB or something which does not require a locus the Demon is incapable of providing.  Doing so requires inter-Demon communication - you are not selling data, you are swapping value.  Fancy EULA's and free trial periods do not burden a Government Demon's workload (at last some good news!).

4) A "fully formed" trade route would have 32 characters (http:/)(/12Origin)/(12Destination) This is the minimum real life information a Demon should have to answer questions about the data in it's possession.

My point is really that the web operates with short cuts which may not always be obvious.  Government requirements will never be identical to Commercial requirements, because neither set is complete.


--- On Tue, 11/9/10, Hugh Barnes <Hugh.Barnes@nehta.gov.au> wrote:

From: Hugh Barnes <Hugh.Barnes@nehta.gov.au>
Subject: RE: Does the user care about URLs? (another thread from Re: Censorship?)
To: rachel.flagg@gsa.gov
Cc: chris@e-beer.net.au, "W3C e-Gov IG" <public-egov-ig@w3.org>
Date: Tuesday, November 9, 2010, 11:18 PM

[ Sent this a few hours back (0201 UTC), thought I had Mike's problem, but seems I trimmed the recipient addresses too keenly. Take two! ]


Great to see you still around, Rachel :)

>    Hi Chris - thanks for the thought-provoking questions below about "does the user care?".... 

I didn't have a clue about a very large chunk of this thread to this point. Hopefully I've been the only one puzzled by its purpose. Now, thanks to Rachel for pointing it out, I can comment on a small part I'd missed.
>    URL/URI structure IS important on the back-end, to help us do a better 
>job managing our information.  A good Info Architecture helps you 
>organize, categorize & manage your information, but I don't think end 
>users care about it

I think *most* users don't *think* they care about it, but they probably care about some of the benefits it brings them unnoticed.

Let's talk about persistent, deterministic URIs:

* visited links show as such (as long as they weren't styled out, grr..)
* caching *should* work
* users know their bookmarks/links/bibliographic citations will continue to work and point to the same resource

But this isn't about persistence and probably very few here doubt its benefits.

It *is* closely related to other good practices, which include the next list of benefits – namely systematic and descriptive URI patterns:

* on the address bar, though most browsers will search the history for page titles these days too, which is handy, URIs provide another hook for address bar discovery
* when shared, descriptive URIs provide strong cues about the content and permit trust (contrast with, e.g. bit.ly)
* URLs appearing in print make that little bit more sense ("I should look that one up when I get back online")
* search engines rightly weight them highly, along with important markup elements
* they can be marketed or read out more easily over the phone, on radio, or on television (pronouncability)
* they have more chance of being remembered (have you ever seen an interesting URL outside but couldn't make a note of it?)
* users can, over a slow period of indoctrination :), start to "get" URIs as structured information paths, not as something "techy" beyond their reach
* conventions like {department}.gov/about or /contact or /publications across government sites, when followed and known, can be hugely beneficial shortcuts (principle of least surprise)

Advanced users (which are not "most" users, but actually an influential group):

* tweak URLs, like version numbers, dates, or query keywords, or adding "/feed"
* dislike system information in the URI, like "joomla" or "/default.aspx" because they know it'll change when the new CMS is implemented and also because they are less likely to remember those details (if they do it's a waste of space!). (That means they're less likely to link or reference a URI, which in turn affects most users.)
* will truncate URLs when they get a 404 (etc.) for some reason – we love it when this comes together
>    They might care about the domain that comes up in search results -- to verify that the site they want to click on is "trusted" -- but as long as people can go to Google or Bing & find what they are looking for, I don't think they care about URI. 
Yes, good search engines certainly have changed the bookmark and URI recall landscape, but not reduced them to insignificance.

I sense a disturbing antagonism toward good practice URLs these days based on the premise they don't really mean anything to most people (which is conditionally true). It's often also driven (IMHO) in a very narrowly focussed way by microblogging. There's some truth to it, but it's just not that simple. I would be sad to see a long-held best architectural practice discarded. I think government resources can and should lead in this respect.

Good discussion BTW, Chris, I look forward to seeing if there's more to say.


Hugh Barnes
Technical Interface Specialist
nehta – National E-Health Transition Authority
Address:      Level 2, 10 Browning St, West End, QLD, 4101
Phone:        +61 7 3023 8537
Mobile:       +61 417 469 552
Email:        hugh.barnes@nehta.gov.au
Web:          http://www.nehta.gov.au

Received on Wednesday, 10 November 2010 19:49:23 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:00:44 UTC