[
Lists Home |
Date Index |
Thread Index
]
Yes, avoiding querystrings when you want the deferenced item to be
accessed by bots, or when the item should be idempotent, would seem a
good best practice.
But a follow-on question might be: when is it a good idea to use
querystrings?
Ian
bryan rasmussen wrote:
> Well I think there is a best-practice scenario for avoiding
> querystrings as much as possible, based simply on the difficulty of
> automated agents, such as search engine bots, of differentiating
> between different content that is made different by the structure of
> its querystrings.
>
> Cheers,
> Bryan Rasmussen
>
> On 12/10/05, Ian Graham <ian.graham@utoronto.ca> wrote:
>
>>You are quite right -- this is very much a human user-centric approach
>>to URLs. Really, then, my example reflects one use case for one type of
>>actor.
>>
>>So, going back to Roger's original question (best practices for URI
>>construction), it is likely true that there is no one set of best
>>practices. The best practices you want will depend on the application
>>you are building: that is, on actors who will be using your URIs, and
>>the use cases that are relevant.
>>
>>Does that make sense?
>>
>>[Aside: don't you like the way Roger can post seemingly simple
>> questions that somehow always lead to incredibly interesting
>> discussions!]
>>
>>Greg Hunt wrote:
>>
>>>Isn't that a specific case of a URL referring to a human-readable web
>>>page in a site addressed to people? When the resource does not provide
>>>for navigation (media files or database files do not usually point at
>>>each other, web pages usually do point at each other) truncating the URL
>>>is an attempt to get to a web page (in essentially a parallel
>>>information structure) that then points to the thing that you want. The
>>>fall-back to web pages is not part of the general resource model and
>>>nothing stops anyone from creating a site that contains no html if they
>>>only want to to provide resources other than web pages.
>>>People do all kinds of bizarre things with web browsers and a
>>>human-readable web site architecture should accommodate that, but web
>>>site architecture and resource identification are different things.
>>>Greg
>>>
>>>Ian Graham wrote:
>>>
>>>
>>>>
>>>>Bjoern Hoehrmann wrote:
>>>>
>>>>
>>>>>* Ian Graham wrote:
>>>>>
>>>>>
>>>>>>One observation/result was that a substantial fraction of users
>>>>>>expect truncating a URL at 'natural' places to return meaningful
>>>>>>results. This, for example, is one of the intuitive 'search'
>>>>>>algorithms people use when a link doesn't work. As an example, if
>>>>>>
>>>>>>http://somewhere.org/US/IL/
>>>>>>
>>>>>>is a valid URL, then truncating this to
>>>>>>
>>>>>>http://somewhere.org/US/
>>>>>>
>>>>>>should both work and provide meaningful information.
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>People who do this commit web architecture crime, the TAG commands
>>>>>"People making use of URIs SHOULD NOT attempt to infer properties
>>>>>of the referenced resource".
>>>>
>>>>
>>>>
>>>>Well, since likely < 0.002% of Web users have read the spec, and since
>>>>that recommendation is counter to any intuitive understanding of a
>>>>URL, that particular recommendation is .... err... useless.
>>>>
>>>>People do what make sense to them. And since URLs have structure, some
>>>>people (and likely some software) naturally look for meaning in it.
>>>>
>>>>I mean how often have you gone to a URL, got a 404, and then chopped
>>>>bits off until you found something relevant/useful? And how often
>>>>were you pissed when that didn't work?
>>>>
>>>>Designing for what people actually do is, imo, a characteristic of
>>>>good design, even if this contradicts a spec.
>>>>
>>>>Ian
>>>
>>>
>>>
>>>-----------------------------------------------------------------
>>>The xml-dev list is sponsored by XML.org <http://www.xml.org>, an
>>>initiative of OASIS <http://www.oasis-open.org>
>>>
>>>The list archives are at http://lists.xml.org/archives/xml-dev/
>>>
>>>To subscribe or unsubscribe from this list use the subscription
>>>manager: <http://www.oasis-open.org/mlmanage/index.php>
>>>
>>--
>>Ian Graham
>>H: 416.769.2422 / W: 416.513.5656 / E: <ian . graham AT utoronto . ca>
>>
>>-----------------------------------------------------------------
>>The xml-dev list is sponsored by XML.org <http://www.xml.org>, an
>>initiative of OASIS <http://www.oasis-open.org>
>>
>>The list archives are at http://lists.xml.org/archives/xml-dev/
>>
>>To subscribe or unsubscribe from this list use the subscription
>>manager: <http://www.oasis-open.org/mlmanage/index.php>
>>
>>
--
Ian Graham
H: 416.769.2422 / W: 416.513.5656 / E: <ian . graham AT utoronto . ca>
|
