CT: URI patterns

[Charles McCathieNevile-2]

There are a couple more patterns that are not currently mentioned in  
appendix E:

http://wap.* and http://m.*

To be honest, I still hate the idea of using URI patterns in a  
specification as general as this one.

cheers

Chaals

--
Charles McCathieNevile  Opera Software, Standards Group
     je parle français -- hablo español -- jeg lærer norsk
http://my.opera.com/chaals       Try Opera: http://www.opera.com

[Phil Archer-4]

If you'll forgive me crossing my boundaries here...

Appendix E assumes that everyone understands what *.mobi means. I agree
with Charles that other patterns are, de facto, associated with mobile.

Without wishing to add to the burden, one could make this idea a little
more extensible and less arbitrary with reference to [1]. This uses some
EBNF written originally by the folk now working on the widgets specs, to
define URI patterns. We use it in POWDER but we didn't write it.

Something like 'following the method for defining URI patterns as given
in POWDER Grouping of Resources, we note the following patterns...'

Phil.

[1] http://www.w3.org/TR/2009/REC-powder-grouping-20090901/#wild


Charles McCathieNevile wrote:

> There are a couple more patterns that are not currently mentioned in
> appendix E:
>
> http://wap.* and http://m.*
>
> To be honest, I still hate the idea of using URI patterns in a
> specification as general as this one.
>
> cheers
>
> Chaals
>

--


Phil Archer
W3C Mobile Web Initiative
http://www.w3.org/Mobile

http://philarcher.org

[Charles McCathieNevile-2]

On Mon, 28 Sep 2009 10:07:59 +0200, Phil Archer <phila@...> wrote:

> If you'll forgive me crossing my boundaries here...
>
> Appendix E assumes that everyone understands what *.mobi means. I agree  
> with Charles that other patterns are, de facto, associated with mobile.
>
> Without wishing to add to the burden, one could make this idea a little  
> more extensible and less arbitrary with reference to [1]. [...]
>
> Something like 'following the method for defining URI patterns as given  
> in POWDER Grouping of Resources, we note the following patterns...'

> [1] http://www.w3.org/TR/2009/REC-powder-grouping-20090901/#wild

Yes, providing a pointer that explains what the syntax means (or even  
copying it) would be a good idea.

cheers

Chaals

--
Charles McCathieNevile  Opera Software, Standards Group
     je parle français -- hablo español -- jeg lærer norsk
http://my.opera.com/chaals       Try Opera: http://www.opera.com

[Jo Rabin-2]

Hi Chaals

We considered other URI patters and they were dropped as part of the
normative/non-normative heuristic discussions under ISSUE-288 and
resolved in March [1].

[1] http://lists.w3.org/Archives/Public/public-bpwg/2009Mar/0174.html

Of course any resolution is open to being torn up and contradicted by a
subsequent resolution, but before doing so I would appreciate detailed
commentary as to why that is necessary given the depth of previous
discussion on the topic.

FWIW as a passionate fan of URI Patterns for Grouping of resources I'd
be happy to use that reference but for a list of 1 pattern it may be a
bit excessive.

Jo

On Mon, 28 Sep 2009 10:07:59 +0200, Phil Archer <phila@...> wrote:

 > If you'll forgive me crossing my boundaries here...
 >
 > Appendix E assumes that everyone understands what *.mobi means. I agree
 > with Charles that other patterns are, de facto, associated with mobile.
 >
 > Without wishing to add to the burden, one could make this idea a little
 > more extensible and less arbitrary with reference to [1]. [...]
 >
 > Something like 'following the method for defining URI patterns as given
 > in POWDER Grouping of Resources, we note the following patterns...'

 > [1] http://www.w3.org/TR/2009/REC-powder-grouping-20090901/#wild

Yes, providing a pointer that explains what the syntax means (or even
copying it) would be a good idea.

cheers

Chaals

--
Charles McCathieNevile  Opera Software, Standards Group
      je parle français -- hablo español -- jeg lærer norsk
http://my.opera.com/chaals       Try Opera: http://www.opera.com

[Charles McCathieNevile-2]

On Mon, 28 Sep 2009 12:32:06 +0200, Jo Rabin <jrabin@...> wrote:

> We considered other URI patters and they were dropped as part of the  
> normative/non-normative heuristic discussions under ISSUE-288 and  
> resolved in March [1].
>
> [1] http://lists.w3.org/Archives/Public/public-bpwg/2009Mar/0174.html
>
> Of course any resolution is open to being torn up and contradicted by a  
> subsequent resolution, but before doing so I would appreciate detailed  
> commentary as to why that is necessary given the depth of previous  
> discussion on the topic.

Ah. Because I disagree with the resolution, of course :) (I can live with  
it though, so there is no change to the consensus).

> FWIW as a passionate fan of URI Patterns for Grouping of resources I'd  
> be happy to use that reference but for a list of 1 pattern it may be a  
> bit excessive.

While I can live with the group's assertion that there is currently one  
pattern (and even the implication that this list may not be maintained in  
the long term) I think that the pattern used should be more clearly  
specified. If the group is prepared to go with 'anything in the top level  
domain ".mobi" and nothing else' (as far as URI patterns go) then it  
should say so ina  way more easily comprehensible to people. If it wants  
to maintain the idea that another dozen or 50 domains might spring up  
(.mobi.no for example) then a reference to the pattern might be a  
convenient shorthand...

cheers

Chaals

> Jo
>
> On Mon, 28 Sep 2009 10:07:59 +0200, Phil Archer <phila@...> wrote:
>  > Appendix E assumes that everyone understands what *.mobi means...
>  > Without wishing to add to the burden, one could make this idea a  >  
> little more extensible and less arbitrary with reference to [1].
>  > [...]
>  > [1] http://www.w3.org/TR/2009/REC-powder-grouping-20090901/#wild
>
> Yes, providing a pointer that explains what the syntax means (or even
> copying it) would be a good idea.

--
Charles McCathieNevile  Opera Software, Standards Group
     je parle français -- hablo español -- jeg lærer norsk
http://my.opera.com/chaals       Try Opera: http://www.opera.com