>...I didn't understand the proposal
>assumed additional requirements for URL/URI parsers, so I didn't
>understand its usefulness. **If we can allow that**, I see this can
>be useful, while it should be minor usage ...
Certainly, it's envisioned to be a small niche, which is why I am
not ready t
> On Fri, 8 Jul 2005 12:16:48 -0400,
> Bill Fenner <[EMAIL PROTECTED]> said:
>> So, I guess the appropriate next step for this work is to make
>> consensus on this, which mostly equals to my question -1:
>>
>> -1. are we okay with forcing URL/URI parsers to understand the
>> detailed sem
>So, I guess the appropriate next step for this work is to make
>consensus on this, which mostly equals to my question -1:
>
> -1. are we okay with forcing URL/URI parsers to understand the
> detailed semantics of the scoped address syntax and to strip the
> zone ID (+ delimiter) part b
>
> > I don't understand this use case. Assuming I have a router
> > and it's manual says type in
> > http://de0/
> >
>
I guess in that case that would be a doc error from
that vendor. I would expect the doc to say something like
http://<"name of your nic">
>
> A ve
>
> So, I guess the appropriate next step for this work is to
> make consensus on this, which mostly equals to my question -1:
>
> -1. are we okay with forcing URL/URI parsers to understand the
> detailed semantics of the scoped address syntax and to strip the
> zone ID (+ delimit
(Just coming back to normal work from a vacation, sorry for the
delayed response)
> On Mon, 4 Apr 2005 07:37:45 -0800,
> Bill Fenner <[EMAIL PROTECTED]> said:
>> Is my understanding now correct?
> Yes, that looks right. And even if getaddrinfo took whatever
> form directly (either the
Perry Lorier <[EMAIL PROTECTED]> writes:
> Bill Fenner wrote:
> > I usually think of the small home router configuration problem -
> > buy a box, plug it in, it wants you to configure it using a web
> > page, and it's probably fe80::1. I don't have any systems in my
> > house that have fewer than
--kc
-Original Message-
From:
JINMEI Tatuya / <[EMAIL PROTECTED]>
Sent: Apr 4, 2005 12:33 AM
To: Bill Fenner
Cc: ipv6@ietf.org
Subject: Re: Move forward with scoped literal URI format?
2. then the browser parser parses the entire URL and extracts
"v6.fe80::1_de0" and
On Mon, 4 Apr 2005, JINMEI Tatuya / [ISO-2022-JP] ¿ÀÌÀãºÈ wrote:
1. assume we type "http://[v6.fe80::1_de0]/"; in "the URL bar" of the
browser.
I doubt hardly any parsers accept this "v6." notation, and I'd rather
they even wouldn't.
Best just to forget about the whole thing? :)
--
Pekka Savol
On Mon, 04 Apr 2005 16:33:00 +0900,
JINMEI Tatuya <[EMAIL PROTECTED]> said:
>Is my understanding now correct?
Yes, that looks right. And even if getaddrinfo took whatever
form directly (either the separator is '%' or getaddrinfo is
modified to accept the URI character as well), I think it's
reas
Bill Fenner wrote:
> I usually think of the small home router configuration problem -
> buy a box, plug it in, it wants you to configure it using a web
> page, and it's probably fe80::1. I don't have any systems in my
> house that have fewer than two non-loopback interfaces. Since
> this is presu
> On Sun, 3 Apr 2005 19:40:36 -0800,
> Bill Fenner said:
>> I guess I don't understand the latter sentence...
>>
>> - what is "this format"? (perhaps it's "fe80::1(some delimiter)de0"
>> for URI)
> It's [v6.fe800:1(some delimiter)de0].
Ah, I see. Then I've been misunderstanding the p
> On Sun, 3 Apr 2005 11:17:17 -0800,
> Bill Fenner said:
>> The essential point is, at least to me, is that we did not want to
>> force applications (like URI/URL parsers) to be aware of scope zones
>> and/or the dedicated syntax for scoped addresses.
> My reading was that we don't want
>> On Sun, 3 Apr 2005 11:17:17 -0800,
>> Bill Fenner said:
>> Since this format is unique and is only used for scoped
>> addresses, the application doesn't have to decide based on the address -
>> it's already been told based on the URI format.
>
>I guess I don't understand the latter se
>The essential point is, at least to me, is that we did not want to
>force applications (like URI/URL parsers) to be aware of scope zones
>and/or the dedicated syntax for scoped addresses.
My reading was that we don't want applications to have to examine an
arbitrary address and decide whether or
> > square bracket does not fit the RFC3986 abnf anyways. therefore,
> > i do not think addition of "v6." or use of "_" would really help.
> Please look again at the IP-Literal and IPvFuture productions.
>
> > i would say we should stick to current
> > http://[fe80::1%fxp0
On Wed, Mar 30, 2005 at 06:53:12AM -0800, Bill Fenner wrote:
> > i would say we should stick to current
> > http://[fe80::1%fxp0]:80/index.html
>
> This is not the current notation, neither the grammar in rfc 2732 nor rfc
> 3986 permits it, and rfc 3986 explicitly mentions zones a
> On Wed, 30 Mar 2005 09:18:15 -0800,
> Bill Fenner said:
>> 3. the parser passes "fe80::1_de0" to getaddrinfo(), and gets a
>> sockaddr_in6 structure (whose sin6_addr member is "fe80::1" and
>> sin6_scope_id member is the link ID corresponding to interface
>> "de0"). The browser uses t
, March 30, 2005 11:57 AM
(BTo: Bill Fenner
(BCc: ipv6@ietf.org
(BSubject: Re: Move forward with scoped literal URI format?
(B
(B
(B>>>>> On Wed, 30 Mar 2005 06:56:15 -0800,
(B>>>>> Bill Fenner said:
(B
(B>> Then the browser (parser) implementation would f
You're right, we were out of sync;
>3. the parser passes "fe80::1_de0" to getaddrinfo(), and gets a
> sockaddr_in6 structure (whose sin6_addr member is "fe80::1" and
> sin6_scope_id member is the link ID corresponding to interface
> "de0"). The browser uses the sockaddr_in6 structure with
> On Wed, 30 Mar 2005 06:56:15 -0800,
> Bill Fenner said:
>> Then the browser (parser) implementation would first extract
>> "fe80::1_de0" and pass it to getaddrinfo(3) for converting it to an
>> IPv6 address. So far, so good, but then the browser would also need
>> to modify the entire
>Then the browser (parser) implementation would first extract
>"fe80::1_de0" and pass it to getaddrinfo(3) for converting it to an
>IPv6 address. So far, so good, but then the browser would also need
>to modify the entire URL to:
>
> http://[fe80::1]/
>
>before sending it to the web server on th
> square bracket does not fit the RFC3986 abnf anyways. therefore,
> i do not think addition of "v6." or use of "_" would really help.
Please look again at the IP-Literal and IPvFuture productions.
> i would say we should stick to current
> http://[fe80::1%fxp0]:
> At the IETF meeting in Minneapolis, I talked about the URI format for
> scoped addresses, described in
> http://www.ietf.org/internet-drafts/draft-fenner-literal-zone-01.txt
square bracket does not fit the RFC3986 abnf anyways. therefore,
i do not think addition of "v6." or u
Mukesh.K.Gupta Wrote:
About the character used, I was in favor of "_" before I
read Brian's comment about "_" being vanishing when the
URI is underlined. Considering the number of places where
URIs are underlined (word underlines the URIs, web pages
underline the URIs etc), I don't think using "_
> On Tue, 29 Mar 2005 11:26:01 -0800,
> "Li, Qing" <[EMAIL PROTECTED]> said:
>> 0. Should we solve this problem at all?
> Yes.
> Our boxes are shipped with a console port but the preferred
> manual configuration method is through the browser. The
> preference is to plug the dev
> On Tue, 29 Mar 2005 07:46:01 -0800,
> Bill Fenner said:
> I usually think of the small home router configuration problem -
> buy a box, plug it in, it wants you to configure it using a web
> page, and it's probably fe80::1.
With this type of usage, we would type, e.g.,
http://[fe8
"0. Yes, 1. Yes, 2. No, 3. No"
About the character used, I was in favor of "_" before I
read Brian's comment about "_" being vanishing when the
URI is underlined. Considering the number of places where
URIs are underlined (word underlines the URIs, web pages
underline the URIs etc), I don't think
On Tue, Mar 29, 2005 at 07:22:40PM +0100, Zefram wrote:
> I specifically reject the cut&paste argument in favour
> of using unencoded "%": this is a sufficiently rare situation that
> convenience really doesn't matter.
Users are extremly unlikely to appreciate the fact that non-global
IPv6 addre
On Mon, Mar 28, 2005 at 06:20:19PM -0800, Bill Fenner wrote:
>http://[v6.fe80::cafe:f00d_de0]/ .
Isn't using "v6." here a bit misleading? RFC 3986 seems to say that
the version flag doesn't indicate the IP version, it incidates the
version of the literal format that follows.
David.
-
>
> 0. Should we solve this problem at all?
>
Yes.
Our boxes are shipped with a console port but the preferred
manual configuration method is through the browser. The
preference is to plug the device onto the network and open
the browser and type
http://[v6.fe80::cafe:f00d???fxp
In article <[EMAIL PROTECTED]> (at Tue, 29 Mar 2005 19:22:40 +0100), Zefram
<[EMAIL PROTECTED]> says:
> On the choice of innocuous character I have no strong opinion. "_"
> seems fine. I note that ";" would be harmonious with existing syntax
I disagree to use "_" here because [v6.fe80::1_de0]
Bill Fenner wrote:
>Any other input?
I agree with your analysis: proceed using "_" or some other innocuous
character; do not do anything that requires a change to the established
URI syntax. I specifically reject the cut&paste argument in favour
of using unencoded "%": this is a sufficiently rare
I usually think of the small home router configuration problem -
buy a box, plug it in, it wants you to configure it using a web
page, and it's probably fe80::1. I don't have any systems in my
house that have fewer than two non-loopback interfaces. Since
this is presumably a one-off, I guess the
Juergen Schoenwaelder wrote:
On Mon, Mar 28, 2005 at 06:20:19PM -0800, Bill Fenner wrote:
0. Should we solve this problem at all?
[...]
1. Should we proceed using "_" (or some other non-percent character)?
[...]
2. If not, should we proceed using "%25"?
[...]
3. If not, should we pro
> On Tue, 29 Mar 2005 16:03:11 +0900,
> JINMEI Tatuya <[EMAIL PROTECTED]> said:
>> If the answer is yes, then the question of a delimiter comes up.
>> Percent, as the scoping architecture uses, is problematic because
>> percent is such a special character in URIs.
> I've not yet convince
On Mon, Mar 28, 2005 at 06:20:19PM -0800, Bill Fenner wrote:
> 0. Should we solve this problem at all?
[...]
> 1. Should we proceed using "_" (or some other non-percent character)?
[...]
> 2. If not, should we proceed using "%25"?
[...]
> 3. If not, should we proceed using "%"?
[...]
> On Mon, 28 Mar 2005 18:20:19 -0800,
> Bill Fenner said:
> 0. Should we solve this problem at all?
>The problem is of reaching [fe80::cafe:f00d] via a URI from a
>system attached to multiple links. (note that loopback counts
>as a link on some implementations.) The URI li
38 matches
Mail list logo
