(12:35:43) the Wife: listen (read) these heuristics I found about URLs
(12:35:55) the Wife: Non-confusing links and URLs 
 
    * Links and URLs are unambiguous, clear, and specific, and are as brief as possible.  
 
    * URLs contain human-readable directory and file names that reflect the nature of the information space.  
 
    * URLs are in all lower case and include no special characters, in order to reduce the possibility of typos.  
 
    * Users can tell by reading the text of a link where they will go when they click on it.
(12:36:37) Dave Allen Barker Jr: heh.  Give me a sec. :^)
(12:36:45) Dave Allen Barker Jr: (where did you get that from ?)
(12:38:55) the Wife: It's off of a professional usability person's web page which is off of the Portland Macintosh Users Group
(12:52:52) Dave Allen Barker Jr: I can't find any great quotes to give you on the matter (of URI naming), but I'll pass on a few points I ran across.  Most from the W3C (of course! ;^)).
(12:54:01) Dave Allen Barker Jr: From TimBL's "Cool URIs don't change.":

Topics and Classification by subject  I'll go into this danger in more detail as it is one of the more difficult things to avoid. Typically, topics end up in URIs when you classify your documents according to a breakdown of the work you are doing. That breakdown will change. Names for areas will change. At W3C we wanted to change "MarkUp" to "Markup" and then to "HTML" to reflect the actual content of the section. Also, beware that this is often a flat name space. In 100 years are you sure you won't want to reuse anything? We wanted to reuse "History" and "Stylesheets" for example in our short life.

This is a tempting way of organizing a web site - and indeed a tempting way of organizing anything, including the whole web. It is a great medium term solution but has serious drawbacks in the long term

Part of the reasons for this lie in the philosophy of meaning. every term in the language it a potential clustering subject, and each person can have a different idea of what it means. Because the relationships between subjects are web-like rather than tree-like, even for people who agree on a web may pick a different tree representation. These are my (oft repeated) general comments on the dangers of hierarchical classification as a general solution.

Effectively, when you use a topic name in a URI you are binding yourself to some classification. You may in the future prefer a different one. Then, the URI will be liable to break.

A reason for using a topic area as part of the URI is that responsibility for sub-parts of a URI space is typically delegated, and then you need a name for the organizational body - the subdivision or group or whatever - which has responsibility for that sub-space. This is binding your URIs to the organizational structure. It is typically safe only when protected by a date further up the URI (to the left of it): 1998/pics can be taken to mean for your server "what we meant in 1998 by pics", rather than "what in 1998 we did with what we now refer to as pics."

(12:58:34) the Wife: The guy makes interesting points about URI names...so if he doesn't like hierarchies...how about everything in one directory with a relatively informative name.
(12:59:42) Dave Allen Barker Jr: That guy is the founder of the Web and the W3C (working on the semantic web too), so he's thought about this about.
(13:00:37) Dave Allen Barker Jr: I don't think he's saying go whole hog abstract (like I've thought about in the past), but is cautioning people about the warm and fuzzy feeling they get when they think up a perfect name (URL) for things.
(13:01:16) the Wife: URIs being numbers may sound like a solution, however the author is going to have to understand the clustering/grouping/hierarchy of the files, and therefore the ideas, just to find a file.  And what about maintaining these files?  You would have to build some sort of app that searches through all the html files so you can find the appropriate one to edit...
(13:01:26) Dave Allen Barker Jr: I completely understand the view and motivation behind the suggestions of the person you passed along.  I just think it's a bit simple or near sighted.
(13:02:31) the Wife: I agree to some degree
(13:02:50) the Wife: Just thought I would share
(13:04:00) the Wife: When you have a website who's sole intention is to dispense information, you want to set up groupinsg and a heirarchy as a way of explaining the info to the user.  As a way of aiding in the formation of a correct mental model
(13:04:39) Dave Allen Barker Jr: Only if he has to find the files by filename.  He shouldn't have to.  It's the same reason we use varables intead of memory locations, and actually, filenames instead of inodes on disk.  Yeah, it can be hard to do with todays convetional tools (filebrowsers).  But more and more people are getting the hang of it (internet search engines, labels on email messages in readers, etc.).
(13:05:02) Dave Allen Barker Jr: For sure, but you don't want to "hardwire" that into the URL.
(13:05:51) Dave Allen Barker Jr: It's be nice to have per user hierarchies organizing the same resources (identified by URLs) in customized ways.
(13:06:37) the Wife: I don't know if a user would want to take the time to organize info.  
(13:06:46) Dave Allen Barker Jr: That's what the UI is for, and a URL only has a limited place as a UI element in my mind. (homepage for sure, maybe a couple others).
(13:07:01) Dave Allen Barker Jr: No, the user wouldn't, you would as an Information Architect.
(13:07:28) the Wife: Well the URI doesn't have to refelct the web site's hierarchy
(13:07:58) the Wife: It just has to be relevantly named to the page's purpose
(13:08:02) Dave Allen Barker Jr: *exactly*  Isn't that contrary to the opinion of the person you passed on?
(13:08:57) Dave Allen Barker Jr: That's where I disagree.  Relavence changes, even when the value of the resource (URL) itself does not.
(13:09:09) the Wife: URLs contain human-readable directory and file names that reflect the nature of the information space.
(13:09:38) the Wife: *the nature* of the info space...not the architecture or hierarchy or structure
(13:09:48) the Wife: I read nature as purpose
(13:10:30) Dave Allen Barker Jr: they don't have to.  They just did for a while because simple webserver configurations use to reflect filesystems, and filesystems were a simple way to organize data (as a heirarchy).
(13:10:34) the Wife: If you have a FAQs webpage, you may call it FAQ, not a name describing the most frequently asked questions at that time
(13:11:35) Dave Allen Barker Jr: I didn't follow the point of your last line..
(13:11:58) the Wife: The purpose of the FAQ webpage is to display and answer the FAQs
(13:12:02) the Wife: Call it FAQ
(13:12:26) the Wife: or blah.blah/faq.html
(13:12:50) the Wife: don't call it based on the data contained inside the webpage (the questions and answers)
(13:12:58) Dave Allen Barker Jr: that's nice, and it just might work.  But that's a pretty small naming challange.
(13:13:17) the Wife: 100 years from now there still may be a FAQ webpage (blah.blah/faq.html), but the info on it will change
(13:14:10) Dave Allen Barker Jr: Are you trying to teach me something you don't think I get, you just showing me you know something (I'm *not* trying to be "snippy"! :^))?
(13:17:19) the Wife: Well there was no way I could respond to your last remark
(13:20:21) the Wife: I have no problem with the lack of (or at least reduced) hierarchy in the URI.  But I also think it is ok to use it as a way of organizing the tremendous amt of info available to users
(13:21:37) the Wife: If a person puts in a URI that no longer exists, it can be redirected to a home page that displays the current hierarchy
(13:21:44) the Wife: Big whoop
(13:21:49) Dave Allen Barker Jr: If that were the case, we wouldn't provide navigation in the page.  The ease of the heirarchy and the quality of the words chosen in the URL would be so obvious people could just modify the URL like they would construct a command line request.
(13:22:36) Dave Allen Barker Jr: URLs just aren't that handy in that way (beyond a minimally complex organization).
(13:22:57) the Wife: I've done that before...trying to guess where to go...and it sometimes works
(13:22:58) Dave Allen Barker Jr: And yeah, I've been thinking about the impact of redirecting too, wondering if it is a "big whoop".
(13:23:45) Dave Allen Barker Jr: But didn't you recognize it as a hack?  Like the site was broken?
(13:24:01) the Wife: But if I have someone emailing me a URL, or if there is a link on a page, I can somewhat guess what will be on the page based on the URI
(13:24:16) the Wife: I would be more likely to visit (or not visit) it.
(13:24:59) the Wife: No...I saw it as an additional way to navigate a website organized in a way I could understand.
(13:25:36) the Wife: And your comment assumed that just because the URI's name provided info that the web site itself didn't???
(13:25:56) Dave Allen Barker Jr: Only because you have to (read the URL).  Links are better passed around as titles to where they point (Dave's Homepage).  You see filenames and not inodes on the disk, you see titles in your list of bookmarks, not URLs, etc...
(13:26:26) the Wife: right, but when you hover over the link, you can then see the URL
(13:26:32) the Wife: I like having both
(13:27:34) the Wife: You use the info from a URL all the time...you found out that a Pug Owner's web page is in Japan even though their local number is Atlanta
(13:27:38) Dave Allen Barker Jr: I do too!  Becaues the web is still imature and built by people getting the hang of it.  I'd argue that a well done website should be completely usable without a status bar (where URLs are displayed).
(13:27:56) the Wife: I found out that this guy who wrote the URL rules was part of the Portland Mac Users Group
(13:28:21) the Wife: I wouldn't trust it
(13:28:25) Dave Allen Barker Jr: You should check out the URLs you're clicking on in Ebay.
(13:29:11) the Wife: Of course there are different ways to do it...but I wouldn't want my personal or professional web address to be non-sensical
(13:30:23) Dave Allen Barker Jr: (about the pug page): You're right, I did, but I don't think it was anything the webmaster was trying to communicate.  Any more that me looking up who own the IP addresses of a site, or the home phone number of an author,  Yeah, there are many channels to more information from the place I started, but the weren't neccessary to use the site as intended.
(13:31:16) Dave Allen Barker Jr: Like "google"? ;^)
(13:31:19) the Wife: well, then we can agree to disagree
(13:32:16) the Wife: Huh?  I don't understand the google point
(13:32:36) Dave Allen Barker Jr: I don't think we are (disagreeing).  But yeah, we can drop it for now.  I think the real difference in where we're coming from is not that decisive.
(13:34:04) Dave Allen Barker Jr: Google is an absolutely nonsensical way to access a search engine service.  But all those URLs had been taken (find.com, search.com, etc.).  It's no different that 0ne.org in representing the hosted information, except that they have an effective marketing machine.
(13:35:17) the Wife: That's a different discussion...that is about a company's name and identity.  Not the fact that your search criteria is embedded in a URL name
(13:35:21) Dave Allen Barker Jr: Not that I don't, my marketing machine just doesn't have to be as invovled because the people I'm marketing too have me bookmarked after I told them about me.
(13:35:35) the Wife: You would hope
(13:35:53) the Wife: But if I get on a new computer or use a new browser...I no longer have the bookmarks
(13:36:11) Dave Allen Barker Jr: That's what google is for! ;^)
(13:37:39) Dave Allen Barker Jr: (Or a find command for a filesystem, or a phonebook for a phone number, etc.)