fb: URIs?

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
37 messages Options
12
Reply | Threaded
Open this post in threaded view
|

fb: URIs?

Daniel R. Tobias
I just noticed that the iPhone Facebook app, when you enable its
recently-added feature to sync your Facebook friends with your iPhone
contacts, inserts URIs for each person of the form:

fb://profile/771025267

When you click on such a URI from the iPhone contact section, it
brings up the person's Facebook info via the Facebook app.

I presume this is a nonstandard, unregistered URI scheme; has any
attempt been made to register it?

It also appears to abuse the double slash, since what follows doesn't
seem to be any sort of "authority".

When the contacts are further synced to other programs and systems (I
have mine automatically syncing in quite a few directions to various
things both on my PC and on the net), you end up with nonfunctional
links in most of the places, as no programs that I know of outside
the iPhone support this scheme; the "use HTTP for everything" crowd
sometimes has a point.


--
== Dan ==
Dan's Mail Format Site: http://mailformat.dan.info/
Dan's Web Tips: http://webtips.dan.info/
Dan's Domain Site: http://domains.dan.info/






Reply | Threaded
Open this post in threaded view
|

Re: fb: URIs?

Ira McDonald
Hi Dan,

You're right - 'fb:' is NOT registered and due to
the syntax (invalid authority) COULD NOT be
registered with IANA.

Cheers,
- Ira

Ira McDonald (Musician / Software Architect)
Chair - Linux Foundation Open Printing WG
Co-Chair - TCG Hardcopy WG
IETF Designated Expert - IPP & Printer MIB
Blue Roof Music/High North Inc
email: [hidden email]
winter:
  579 Park Place  Saline, MI  48176
  734-944-0094
summer:
  PO Box 221  Grand Marais, MI 49839
  906-494-2434



On Mon, Feb 15, 2010 at 7:38 AM, Daniel R. Tobias <[hidden email]> wrote:

> I just noticed that the iPhone Facebook app, when you enable its
> recently-added feature to sync your Facebook friends with your iPhone
> contacts, inserts URIs for each person of the form:
>
> fb://profile/771025267
>
> When you click on such a URI from the iPhone contact section, it
> brings up the person's Facebook info via the Facebook app.
>
> I presume this is a nonstandard, unregistered URI scheme; has any
> attempt been made to register it?
>
> It also appears to abuse the double slash, since what follows doesn't
> seem to be any sort of "authority".
>
> When the contacts are further synced to other programs and systems (I
> have mine automatically syncing in quite a few directions to various
> things both on my PC and on the net), you end up with nonfunctional
> links in most of the places, as no programs that I know of outside
> the iPhone support this scheme; the "use HTTP for everything" crowd
> sometimes has a point.
>
>
> --
> == Dan ==
> Dan's Mail Format Site: http://mailformat.dan.info/
> Dan's Web Tips: http://webtips.dan.info/
> Dan's Domain Site: http://domains.dan.info/
>
>
>
>
>
>
>

Reply | Threaded
Open this post in threaded view
|

Re: fb: URIs?

Gannon Dick
Ira,

If what Dan is pointing out is correct, then the US Library of Congress "Handle Server" (for the Code lists) is a problem too, I think.  The INFO domain is registered with IANA, but the URL's are of the form

e.g. info:lc/vocabulary/countries/af (for "Afghanistan")

http://www.loc.gov/standards/codelists/

Would you or Dan be so kind as to give me a link to the spec?  Shortening the URL would be very handy.  The straight HTTP/LDAP approach is a bit cumbersome off to the right of the links:
http://www.rustprivacy.org/meta/strat-ml/iso3166rdf.html

--Gannon

--- On Mon, 2/15/10, Ira McDonald <[hidden email]> wrote:

> From: Ira McDonald <[hidden email]>
> Subject: Re: fb: URIs?
> To: "Daniel R. Tobias" <[hidden email]>, "Ira McDonald" <[hidden email]>
> Cc: [hidden email]
> Date: Monday, February 15, 2010, 11:33 AM
> Hi Dan,
>
> You're right - 'fb:' is NOT registered and due to
> the syntax (invalid authority) COULD NOT be
> registered with IANA.
>
> Cheers,
> - Ira
>
> Ira McDonald (Musician / Software Architect)
> Chair - Linux Foundation Open Printing WG
> Co-Chair - TCG Hardcopy WG
> IETF Designated Expert - IPP & Printer MIB
> Blue Roof Music/High North Inc
> email: [hidden email]
> winter:
>   579 Park Place  Saline, MI  48176
>   734-944-0094
> summer:
>   PO Box 221  Grand Marais, MI 49839
>   906-494-2434
>
>
>
> On Mon, Feb 15, 2010 at 7:38 AM, Daniel R. Tobias <[hidden email]>
> wrote:
> > I just noticed that the iPhone Facebook app, when you
> enable its
> > recently-added feature to sync your Facebook friends
> with your iPhone
> > contacts, inserts URIs for each person of the form:
> >
> > fb://profile/771025267
> >
> > When you click on such a URI from the iPhone contact
> section, it
> > brings up the person's Facebook info via the Facebook
> app.
> >
> > I presume this is a nonstandard, unregistered URI
> scheme; has any
> > attempt been made to register it?
> >
> > It also appears to abuse the double slash, since what
> follows doesn't
> > seem to be any sort of "authority".
> >
> > When the contacts are further synced to other programs
> and systems (I
> > have mine automatically syncing in quite a few
> directions to various
> > things both on my PC and on the net), you end up with
> nonfunctional
> > links in most of the places, as no programs that I
> know of outside
> > the iPhone support this scheme; the "use HTTP for
> everything" crowd
> > sometimes has a point.
> >
> >
> > --
> > == Dan ==
> > Dan's Mail Format Site: http://mailformat.dan.info/
> > Dan's Web Tips: http://webtips.dan.info/
> > Dan's Domain Site: http://domains.dan.info/
> >
> >
> >
> >
> >
> >
> >
>
>




Reply | Threaded
Open this post in threaded view
|

Re: fb: URIs?

Ira McDonald
Hi Gannon,

The two specs for URI syntax are:

ftp://ftp.ietf.org/rfc/rfc3986.txt (IETF Standard)
- Uniform Resource Identifier (URI): Generic Syntax

ftp://ftp.ietf.org/rfc/rfc3987.txt (IETF Proposed Standard)
- Internationalized Resource Identifiers (IRIs)

The spec for URI scheme registration rules is:

ftp://ftp.ietf.org/rfc/rfc4395.txt (IETF Best Current Practice)
Guidelines and Registration Procedures for New URI Schemes

The spec for ABNF syntax (used in URI scheme definitions) is:

ftp://ftp.ietf.org/rfc/rfc5234.txt (IETF Standard)
Augmented BNF for Syntax Specifications: ABNF

The idea with RFC 3986 is that you import and use verbatim
as many elements as possible, for reliable processing.

Cheers,
- Ira

Ira McDonald (Musician / Software Architect)
Chair - Linux Foundation Open Printing WG
Co-Chair - TCG Hardcopy WG
IETF Designated Expert - IPP & Printer MIB
Blue Roof Music/High North Inc
email: [hidden email]
winter:
  579 Park Place  Saline, MI  48176
  734-944-0094
summer:
  PO Box 221  Grand Marais, MI 49839
  906-494-2434



On Mon, Feb 15, 2010 at 2:46 PM, Gannon Dick <[hidden email]> wrote:

> Ira,
>
> If what Dan is pointing out is correct, then the US Library of Congress "Handle Server" (for the Code lists) is a problem too, I think.  The INFO domain is registered with IANA, but the URL's are of the form
>
> e.g. info:lc/vocabulary/countries/af (for "Afghanistan")
>
> http://www.loc.gov/standards/codelists/
>
> Would you or Dan be so kind as to give me a link to the spec?  Shortening the URL would be very handy.  The straight HTTP/LDAP approach is a bit cumbersome off to the right of the links:
> http://www.rustprivacy.org/meta/strat-ml/iso3166rdf.html
>
> --Gannon
>
> --- On Mon, 2/15/10, Ira McDonald <[hidden email]> wrote:
>
>> From: Ira McDonald <[hidden email]>
>> Subject: Re: fb: URIs?
>> To: "Daniel R. Tobias" <[hidden email]>, "Ira McDonald" <[hidden email]>
>> Cc: [hidden email]
>> Date: Monday, February 15, 2010, 11:33 AM
>> Hi Dan,
>>
>> You're right - 'fb:' is NOT registered and due to
>> the syntax (invalid authority) COULD NOT be
>> registered with IANA.
>>
>> Cheers,
>> - Ira
>>
>> Ira McDonald (Musician / Software Architect)
>> Chair - Linux Foundation Open Printing WG
>> Co-Chair - TCG Hardcopy WG
>> IETF Designated Expert - IPP & Printer MIB
>> Blue Roof Music/High North Inc
>> email: [hidden email]
>> winter:
>>   579 Park Place  Saline, MI  48176
>>   734-944-0094
>> summer:
>>   PO Box 221  Grand Marais, MI 49839
>>   906-494-2434
>>
>>
>>
>> On Mon, Feb 15, 2010 at 7:38 AM, Daniel R. Tobias <[hidden email]>
>> wrote:
>> > I just noticed that the iPhone Facebook app, when you
>> enable its
>> > recently-added feature to sync your Facebook friends
>> with your iPhone
>> > contacts, inserts URIs for each person of the form:
>> >
>> > fb://profile/771025267
>> >
>> > When you click on such a URI from the iPhone contact
>> section, it
>> > brings up the person's Facebook info via the Facebook
>> app.
>> >
>> > I presume this is a nonstandard, unregistered URI
>> scheme; has any
>> > attempt been made to register it?
>> >
>> > It also appears to abuse the double slash, since what
>> follows doesn't
>> > seem to be any sort of "authority".
>> >
>> > When the contacts are further synced to other programs
>> and systems (I
>> > have mine automatically syncing in quite a few
>> directions to various
>> > things both on my PC and on the net), you end up with
>> nonfunctional
>> > links in most of the places, as no programs that I
>> know of outside
>> > the iPhone support this scheme; the "use HTTP for
>> everything" crowd
>> > sometimes has a point.
>> >
>> >
>> > --
>> > == Dan ==
>> > Dan's Mail Format Site: http://mailformat.dan.info/
>> > Dan's Web Tips: http://webtips.dan.info/
>> > Dan's Domain Site: http://domains.dan.info/
>> >
>> >
>> >
>> >
>> >
>> >
>> >
>>
>>
>
>
>
>

Reply | Threaded
Open this post in threaded view
|

Re: fb: URIs?

Gannon Dick
Thanks so much Ira, It seems like what the LOC had in mind was HTTP in the INFO Domain.  Like <a href="http://www.ex.info#(urn#)+">http://www.ex.info#(urn#)+  This sort of thing comes up with encoded entities all the time.  Whoever thought up fb: grabbed the protocol instead of the pipe (domain).  Cool thought, hopefully a "not in my lifetime result".

> >> > links in most of the places, as no programs
> that I
> >> know of outside
> >> > the iPhone support this scheme; the "use HTTP
> for
> >> everything" crowd
> >> > sometimes has a point.
> >> >
> >> >
> >> > --
> >> > == Dan ==



     

Reply | Threaded
Open this post in threaded view
|

Re: fb: URIs?

Dan Brickley-2
In reply to this post by Daniel R. Tobias
On Mon, Feb 15, 2010 at 1:38 PM, Daniel R. Tobias <[hidden email]> wrote:

> I just noticed that the iPhone Facebook app, when you enable its
> recently-added feature to sync your Facebook friends with your iPhone
> contacts, inserts URIs for each person of the form:
>
> fb://profile/771025267
>
> When you click on such a URI from the iPhone contact section, it
> brings up the person's Facebook info via the Facebook app.
>
> I presume this is a nonstandard, unregistered URI scheme; has any
> attempt been made to register it?
>
> It also appears to abuse the double slash, since what follows doesn't
> seem to be any sort of "authority".
>
> When the contacts are further synced to other programs and systems (I
> have mine automatically syncing in quite a few directions to various
> things both on my PC and on the net), you end up with nonfunctional
> links in most of the places, as no programs that I know of outside
> the iPhone support this scheme; the "use HTTP for everything" crowd
> sometimes has a point.

I got in touch with David Recordon at Facebook (cc:'d). His response
copied below with permission.

Dan

> Hey Dan,
> Asked an engineer on our mobile team and here's what he said.
>
> --David
>
> > The origin of those URLs was entirely pragmatic. The iPhone app handles showing
> > different parts of its UI with an internal URL handler that deals with exactly these URLs.
> > When we developed the sync feature, it was natural to simply allow external callers to
> > direct "fb" URLs to the app, and use the same internal handling mechanism that had
> >  been there all along.
> >
> > These URLs were never designed to be used, or useful, outside of the iPhone.
> >
> > The app is actually capable of handling www URLs (so for example
> > "http://www.facebook.com/profile.php?id=4" causes the same thing to happen as
> >  "fb://profile/4"), but we couldn't use those for the sync feature because there is no way
> >  to tell the system to direct just that specific set of http URLs to the app.
> >
> > Long story short, we (certainly I, anyway) have no intentions to make "fb" a real
> >  URL scheme. It just arose out of convenience.

Reply | Threaded
Open this post in threaded view
|

Re: fb: URIs?

Thomas Fruin
On 16 feb 2010, at 22:39, Dan Brickley wrote:

> On Mon, Feb 15, 2010 at 1:38 PM, Daniel R. Tobias <[hidden email]> wrote:
>> I just noticed that the iPhone Facebook app, when you enable its
>> recently-added feature to sync your Facebook friends with your iPhone
>> contacts, inserts URIs for each person of the form:
>>
>> fb://profile/771025267

Unfortunately the FaceBook app doesn't just _insert_ these types of URI's. It also seems to _replace_ already existing URI's of the type http://www.facebook.com/profilename with the new fb:// pseudo URI's.

This means that due to iPhone/Mac syncing, all such URI's that I had manually added to my Mac's contacts have now been converted into fb:// strings, which are unfortunately not very useful on my computer :-( Hopefully this bug will be fixed.

-- Thomas Fruin


Reply | Threaded
Open this post in threaded view
|

Re: fb: URIs?

Graham Klyne-4
In reply to this post by Dan Brickley-2
 >>> Long story short, we (certainly I, anyway) have no intentions to make "fb"
a >>> real URL scheme. It just arose out of convenience.

Hmmm... so what happens to Facebook's iPhone app when someone validly registers
and deploys (widely) a URI scheme called 'fb:'?

I would hazard a suggestion that at least a provisional registration might be in
order (noting its limited use) to help avoid such a thing from happening without
anyone noticing.

#g
--

Dan Brickley wrote:

> On Mon, Feb 15, 2010 at 1:38 PM, Daniel R. Tobias <[hidden email]> wrote:
>> I just noticed that the iPhone Facebook app, when you enable its
>> recently-added feature to sync your Facebook friends with your iPhone
>> contacts, inserts URIs for each person of the form:
>>
>> fb://profile/771025267
>>
>> When you click on such a URI from the iPhone contact section, it
>> brings up the person's Facebook info via the Facebook app.
>>
>> I presume this is a nonstandard, unregistered URI scheme; has any
>> attempt been made to register it?
>>
>> It also appears to abuse the double slash, since what follows doesn't
>> seem to be any sort of "authority".
>>
>> When the contacts are further synced to other programs and systems (I
>> have mine automatically syncing in quite a few directions to various
>> things both on my PC and on the net), you end up with nonfunctional
>> links in most of the places, as no programs that I know of outside
>> the iPhone support this scheme; the "use HTTP for everything" crowd
>> sometimes has a point.
>
> I got in touch with David Recordon at Facebook (cc:'d). His response
> copied below with permission.
>
> Dan
>
>> Hey Dan,
>> Asked an engineer on our mobile team and here's what he said.
>>
>> --David
>>
>>> The origin of those URLs was entirely pragmatic. The iPhone app handles showing
>>> different parts of its UI with an internal URL handler that deals with exactly these URLs.
>>> When we developed the sync feature, it was natural to simply allow external callers to
>>> direct "fb" URLs to the app, and use the same internal handling mechanism that had
>>>  been there all along.
>>>
>>> These URLs were never designed to be used, or useful, outside of the iPhone.
>>>
>>> The app is actually capable of handling www URLs (so for example
>>> "http://www.facebook.com/profile.php?id=4" causes the same thing to happen as
>>>  "fb://profile/4"), but we couldn't use those for the sync feature because there is no way
>>>  to tell the system to direct just that specific set of http URLs to the app.
>>>
>>> Long story short, we (certainly I, anyway) have no intentions to make "fb" a real
>>>  URL scheme. It just arose out of convenience.
>



Reply | Threaded
Open this post in threaded view
|

Re: fb: URIs?

noah_mendelsohn
In reply to this post by Dan Brickley-2
> > These URLs were never designed to be used, or useful, outside of the
iPhone.

Right, but I presume that if someone else where to register fb: and use
fb: URLs for some other purpose, the phone might confused about the new
ones.  It seems to me that even uses within a limited community tend to
cause these shorts of effects, and therefore should be discouraged.  I do
agree that, once someone has implemented a feature like this on a device,
it's hard to blame applications for using it.

Noah

--------------------------------------
Noah Mendelsohn
IBM Corporation
One Rogers Street
Cambridge, MA 02142
1-617-693-4036
--------------------------------------








Dan Brickley <[hidden email]>
Sent by: [hidden email]
02/16/2010 04:39 PM
 
        To:     [hidden email]
        cc:     David Recordon <[hidden email]>, (bcc: Noah
Mendelsohn/Cambridge/IBM)
        Subject:        Re: fb: URIs?


On Mon, Feb 15, 2010 at 1:38 PM, Daniel R. Tobias <[hidden email]> wrote:

> I just noticed that the iPhone Facebook app, when you enable its
> recently-added feature to sync your Facebook friends with your iPhone
> contacts, inserts URIs for each person of the form:
>
> fb://profile/771025267
>
> When you click on such a URI from the iPhone contact section, it
> brings up the person's Facebook info via the Facebook app.
>
> I presume this is a nonstandard, unregistered URI scheme; has any
> attempt been made to register it?
>
> It also appears to abuse the double slash, since what follows doesn't
> seem to be any sort of "authority".
>
> When the contacts are further synced to other programs and systems (I
> have mine automatically syncing in quite a few directions to various
> things both on my PC and on the net), you end up with nonfunctional
> links in most of the places, as no programs that I know of outside
> the iPhone support this scheme; the "use HTTP for everything" crowd
> sometimes has a point.

I got in touch with David Recordon at Facebook (cc:'d). His response
copied below with permission.

Dan

> Hey Dan,
> Asked an engineer on our mobile team and here's what he said.
>
> --David
>
> > The origin of those URLs was entirely pragmatic. The iPhone app
handles showing
> > different parts of its UI with an internal URL handler that deals with
exactly these URLs.
> > When we developed the sync feature, it was natural to simply allow
external callers to
> > direct "fb" URLs to the app, and use the same internal handling
mechanism that had
> >  been there all along.
> >
> > These URLs were never designed to be used, or useful, outside of the
iPhone.
> >
> > The app is actually capable of handling www URLs (so for example
> > "http://www.facebook.com/profile.php?id=4" causes the same thing to
happen as
> >  "fb://profile/4"), but we couldn't use those for the sync feature
because there is no way
> >  to tell the system to direct just that specific set of http URLs to
the app.
> >
> > Long story short, we (certainly I, anyway) have no intentions to make
"fb" a real
> >  URL scheme. It just arose out of convenience.




Reply | Threaded
Open this post in threaded view
|

Re: fb: URIs?

Ira McDonald
In reply to this post by Thomas Fruin
Hi,

Apparently, the 'fb:' pseudo-URI scheme has escaped seriously
into the visible user environment - the replacement of real 'http:'
URIs with equivalent (???) 'fb:' URIs described below by Thomas
is a spectacular bug.

Cheers,
- Ira

On Tue, Feb 16, 2010 at 5:48 PM, Thomas Fruin <[hidden email]> wrote:

> On 16 feb 2010, at 22:39, Dan Brickley wrote:
>
>> On Mon, Feb 15, 2010 at 1:38 PM, Daniel R. Tobias <[hidden email]> wrote:
>>> I just noticed that the iPhone Facebook app, when you enable its
>>> recently-added feature to sync your Facebook friends with your iPhone
>>> contacts, inserts URIs for each person of the form:
>>>
>>> fb://profile/771025267
>
> Unfortunately the FaceBook app doesn't just _insert_ these types of URI's. It also seems to _replace_ already existing URI's of the type http://www.facebook.com/profilename with the new fb:// pseudo URI's.
>
> This means that due to iPhone/Mac syncing, all such URI's that I had manually added to my Mac's contacts have now been converted into fb:// strings, which are unfortunately not very useful on my computer :-( Hopefully this bug will be fixed.
>
> -- Thomas Fruin
>
>
>

Reply | Threaded
Open this post in threaded view
|

Re: fb: URIs?

Julian Reschke
On 17.02.2010 02:13, Ira McDonald wrote:
> Hi,
>
> Apparently, the 'fb:' pseudo-URI scheme has escaped seriously
> into the visible user environment - the replacement of real 'http:'
> URIs with equivalent (???) 'fb:' URIs described below by Thomas
> is a spectacular bug.
> ...

This is what happens when a platform starts to use URIs for things they
weren't designed for. If I recall correctly, this issue has been
discussed on the TAG mailing list a long time ago (once the iPhone SDK
came out?).

Apple is well known for using URIs for things they arent't designed for,
*and* then (consequently?) not registering them (itms and ical come to
mind).

The real question here is: what can we do to educate them?

Best regards, Julian

Reply | Threaded
Open this post in threaded view
|

Re: fb: URIs?

Julian Reschke
In reply to this post by Graham Klyne-4
On 17.02.2010 00:24, Graham Klyne wrote:

>  >>> Long story short, we (certainly I, anyway) have no intentions to
> make "fb" a >>> real URL scheme. It just arose out of convenience.
>
> Hmmm... so what happens to Facebook's iPhone app when someone validly
> registers and deploys (widely) a URI scheme called 'fb:'?
>
> I would hazard a suggestion that at least a provisional registration
> might be in order (noting its limited use) to help avoid such a thing
> from happening without anyone noticing.
> ...

The problem is much bigger, as far as I recall, minting new URI schemes
for invoking iphone applications is even recommended by the SDK, so
there are probably many more.

See discussion around
<http://lists.w3.org/Archives/Public/www-tag/2008Oct/0002.html>.

Best regards, Julian

Reply | Threaded
Open this post in threaded view
|

Re: fb: URIs?

Graham Klyne-4
In reply to this post by Julian Reschke
Julian Reschke wrote:
> Apple is well known for using URIs for things they arent't designed for,
> *and* then (consequently?) not registering them (itms and ical come to
> mind).
>
> The real question here is: what can we do to educate them?

Maybe someone else makes a provisional registration with a big scary health
warning (e.g. "USE OF THIS PRIVATE URI SCHEME IS CONSIDERED HARMFUL" with a link
to (say) a TAG finding explaining why)?

#g


Reply | Threaded
Open this post in threaded view
|

Re: fb: URIs?

Claus Färber-5
In reply to this post by Julian Reschke
Julian Reschke <[hidden email]> schrieb/wrote:
> On 17.02.2010 02:13, Ira McDonald wrote:
>> Apparently, the 'fb:' pseudo-URI scheme has escaped seriously into
>> the visible user environment - the replacement of real 'http:' URIs
>> with equivalent (???) 'fb:' URIs described below by Thomas is a
>> spectacular bug. ...

> This is what happens when a platform starts to use URIs for things they
> weren't designed for. If I recall correctly, this issue has been
> discussed on the TAG mailing list a long time ago (once the iPhone SDK
> came out?).

> Apple is well known for using URIs for things they arent't designed for,
> *and* then (consequently?) not registering them (itms and ical come to
> mind).

> The real question here is: what can we do to educate them?

What is needed is an alternative to (ab)using URIs. Just insisting that
URIs "weren't desgined for" this, does not help.

The "correct" alternative to "fb" URIs would be: register a URN scheme;
they don't indicate a location, after all. However, this approach has
two problems:

- Operating systems (not only the iPhone OS but also Windows comes to
  mind) are desinged to route URIs to applications based on the URI
  scheme. This is "urn" for all URNs.

- Registering a URN scheme requires an RFC and IESG approval.

The "tag" URI scheme does not solve the first problem, either.
Registering a new URI scheme does not solve the second problem.

What's needed is an ad-hoc namespace for URI schemes. For example, all
names containing an underscore could be reserved for domain-based
schemes. Then, one could simply create URIs like this:

  facebook_com:profile/4

Further, extending URIs to specify what to do with URIs (view, play,
edit, ..., route to application named "example_org") -- let's call these
"URI dispositions" -- would often help, too.

An iTunes URI, for example, could read:

  itunes_com+http://example.com/rss_podcast

or even:

  itunes_com+subscribe+http://example.net/xml_podcast

(Potential meaning: use the "itunes_com" app/verb/protocol with the
"subscribe" app/verb/protocol for the ressource located with the "http"
protocol. If "itunes_com" is an application and "subscribe" is a generic  
verb, that means: tell "itunes_com" to "subscribe" to the URL.)

But you could also do:

  subscribe+http://example.net/xml_podcast

(Potential meaning: use the "subscribe" app/verb/protocol for the
ressource located with the "http" protocol. If "subscribe" is a generic  
verb, that means: tell any application to "subscribe" to the URL.)

Other potential uses:

  rsync+ssh://example.net/path/to/dir
  git+ssh://example.com/%2fsrv/git/repo.git

  edit+http://example.com/path
  emacs_gnu_org+edit+http://example.net/path

  play+http://example.com/video.mp4
  edit+http://example.com/video.mp4

  download+http://example.org/path/to/file.zip

  fax+tel:+1.555.555-1234       (short form: fax)
  sms+tel:+1.555.555-9876       (short form: sms)

  http+tls://example.com        (short form: https)

Claus




Reply | Threaded
Open this post in threaded view
|

Re: fb: URIs?

Erik van der Poel
In reply to this post by Graham Klyne-4
+1

On Wed, Feb 17, 2010 at 1:48 AM, Graham Klyne <[hidden email]> wrote:

> Julian Reschke wrote:
>>
>> Apple is well known for using URIs for things they arent't designed for,
>> *and* then (consequently?) not registering them (itms and ical come to
>> mind).
>>
>> The real question here is: what can we do to educate them?
>
> Maybe someone else makes a provisional registration with a big scary health
> warning (e.g. "USE OF THIS PRIVATE URI SCHEME IS CONSIDERED HARMFUL" with a
> link to (say) a TAG finding explaining why)?
>
> #g
>
>
>

Reply | Threaded
Open this post in threaded view
|

RE: fb: URIs?

Eran Hammer-Lahav
In reply to this post by Julian Reschke
This goes beyond Facebook and the iPhone. Pretty much every operating system at this point uses URI schemes as handles/hooks for invoking applications for handling them. This was done in Windows for a decade so Apple didn't invent anything new.

Minting new URI schemes is very difficult. The process provides an easy excuse to simply ignore it. I am not suggesting that people should, but they clearly are (take a look at the list of unregistered schemes - it's LONG). There was long debate on minting an application URI scheme just for this (such as app:fb:something), but I don't think it went far.

EHL

> -----Original Message-----
> From: [hidden email] [mailto:[hidden email]] On Behalf Of Julian
> Reschke
> Sent: Wednesday, February 17, 2010 1:39 AM
> To: Graham Klyne
> Cc: David Recordon; Dan Brickley; [hidden email]
> Subject: Re: fb: URIs?
>
> On 17.02.2010 00:24, Graham Klyne wrote:
> >  >>> Long story short, we (certainly I, anyway) have no intentions to
> > make "fb" a >>> real URL scheme. It just arose out of convenience.
> >
> > Hmmm... so what happens to Facebook's iPhone app when someone
> validly
> > registers and deploys (widely) a URI scheme called 'fb:'?
> >
> > I would hazard a suggestion that at least a provisional registration
> > might be in order (noting its limited use) to help avoid such a thing
> > from happening without anyone noticing.
> > ...
>
> The problem is much bigger, as far as I recall, minting new URI schemes for
> invoking iphone applications is even recommended by the SDK, so there are
> probably many more.
>
> See discussion around
> <http://lists.w3.org/Archives/Public/www-tag/2008Oct/0002.html>.
>
> Best regards, Julian


Reply | Threaded
Open this post in threaded view
|

RE: fb: URIs?

Eran Hammer-Lahav
In reply to this post by Graham Klyne-4
And put it in the registry no one reads? If you want to stop people from abusing the system, you need to either fix the system or accommodate their use case (i.e. work with them, not against them).

EHL

> -----Original Message-----
> From: [hidden email] [mailto:[hidden email]] On Behalf Of
> Graham Klyne
> Sent: Wednesday, February 17, 2010 1:48 AM
> To: Julian Reschke
> Cc: Ira McDonald; Thomas Fruin; [hidden email]; David Recordon; Dan Brickley
> Subject: Re: fb: URIs?
>
> Julian Reschke wrote:
> > Apple is well known for using URIs for things they arent't designed
> > for,
> > *and* then (consequently?) not registering them (itms and ical come to
> > mind).
> >
> > The real question here is: what can we do to educate them?
>
> Maybe someone else makes a provisional registration with a big scary health
> warning (e.g. "USE OF THIS PRIVATE URI SCHEME IS CONSIDERED HARMFUL"
> with a link to (say) a TAG finding explaining why)?
>
> #g
>


Reply | Threaded
Open this post in threaded view
|

Re: fb: URIs?

Julian Reschke
In reply to this post by Eran Hammer-Lahav
On 17.02.2010 18:07, Eran Hammer-Lahav wrote:
> This goes beyond Facebook and the iPhone. Pretty much every operating system at this point uses URI schemes as handles/hooks for invoking applications for handling them. This was done in Windows for a decade so Apple didn't invent anything new.

That's fine.

What's problematic is if this is the only way to interact between
applications.

> Minting new URI schemes is very difficult. The process provides an easy excuse to simply ignore it. I am not suggesting that people should, but they clearly are (take a look at the list of unregistered schemes - it's LONG). There was long debate on minting an application URI scheme just for this (such as app:fb:something), but I don't think it went far.
> ....

Later on:
> And put it in the registry no one reads? If you want to stop people from abusing the system, you need to either fix the system or accommodate their use case (i.e. work with them, not against them).

What exactly is broken with the system?

Best regards, Julian

Reply | Threaded
Open this post in threaded view
|

Re: fb: URIs?

Sandro Hawke
In reply to this post by Eran Hammer-Lahav

> And put it in the registry no one reads? If you want to stop people
> from abusing the system, you need to either fix the system or
> accommodate their use case (i.e. work with them, not against them).

It looks like the problem is that systems are dispatching on the scheme
name instead of on the longest leading substring.  If they did the
latter, then people could use http: or tag: URIs for these apps.

Specifically, we hear that

  http://www.facebook.com/profile.php?id=4
and
 fb://profile/4

are semantically the same [1].  If the systems everywhere could dispatch
on "http://www.facebook.com/profile.php?id=" as easily as they dispatch
on "fb:", it seems like the technical side of this problem might go
away.

(The social part -- people liking to make up URI schemes for personal
reasons -- would remain.  I doubt that's the real problem, though.)

Has anyone seen systems which dispatch on leading prefixes?  Does anyone
see a problem with that design?  It would be more confusing in the case
where no handler is installed, since it could fall back to another
handler, but I think in general that would be useful.

    -- Sandro

[1] http://lists.w3.org/Archives/Public/uri/2010Feb/0018.html

Reply | Threaded
Open this post in threaded view
|

RE: fb: URIs?

Eran Hammer-Lahav
In reply to this post by Julian Reschke


> -----Original Message-----
> From: Julian Reschke [mailto:[hidden email]]
> Sent: Wednesday, February 17, 2010 9:13 AM

> What exactly is broken with the system?

Getting consensus on a new URI scheme is hard. The registry also doesn't reflect reality (more unregistered schemes than registered). For example, I don't know what is going on with the about: scheme but it seems to take a long time for something that seems simple.

EHL

12