The 'go' URI Scheme for the Common Name Resolution Protocol
A compact URI format for pointing at a common-name lookup, or at one record on one server.
Mealling, M., "The 'go' URI Scheme for the Common Name Resolution Protocol", RFC 3368, DOI 10.17487/RFC3368, August 2002, <https://www.rfc-editor.org/info/rfc3368>.
@misc{rfc3368,
series = {Request for Comments},
number = 3368,
howpublished = {RFC 3368},
publisher = {RFC Editor},
doi = {10.17487/RFC3368},
url = {https://www.rfc-editor.org/info/rfc3368},
author = {M. Mealling},
title = {{The 'go' URI Scheme for the Common Name Resolution Protocol}},
pagetotal = 8,
year = 2002,
month = aug,
} RFC text
Contents
Network Working Group M. Mealling
Request for Comments: 3368 VeriSign, Inc.
Category: Standards Track August 2002
The 'go' URI Scheme for the Common Name Resolution Protocol
Status of this Memo
This document specifies an Internet standards track protocol for the
Internet community, and requests discussion and suggestions for
improvements. Please refer to the current edition of the "Internet
Official Protocol Standards" (STD 1) for the standardization state
and status of this protocol. Distribution of this memo is unlimited.
Copyright Notice
Copyright (C) The Internet Society (2002). All Rights Reserved.
Abstract
This document defines a URI scheme, 'go:' to be used with the Common
Name Resolution Protocol. Specifically it lays out the syntactic
components and how those components are used by URI Resolution to
find the available transports for a CNRP service. Care should be
taken with several of the URI components because, while they may look
like components found in other URI schemes, they often do not act
like them. The "go" scheme has more in common with the location
independent "news" scheme than any other URI scheme.
Table of Contents
1. Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . 2
3. Syntax Rules . . . . . . . . . . . . . . . . . . . . . . . . 3
3.1 General Syntax . . . . . . . . . . . . . . . . . . . . . . . 3
3.2 ABNF Grammar . . . . . . . . . . . . . . . . . . . . . . . . 3
3.3 Special Cases and Default Values . . . . . . . . . . . . . . 4
3.3.1 If There is Only a Server . . . . . . . . . . . . . . . . . 4
3.3.2 If Server is Empty Then server=localhost . . . . . . . . . . 4
3.3.3 Default Port . . . . . . . . . . . . . . . . . . . . . . . . 4
3.4 Encoding Rules . . . . . . . . . . . . . . . . . . . . . . . 4
4. Transport Independence . . . . . . . . . . . . . . . . . . . 4
5. Examples . . . . . . . . . . . . . . . . . . . . . . . . . . 4
6. Security Considerations . . . . . . . . . . . . . . . . . . 5
7. IANA Considerations . . . . . . . . . . . . . . . . . . . . 5
References . . . . . . . . . . . . . . . . . . . . . . . . . 6
A. Registration Template . . . . . . . . . . . . . . . . . . . 7
Author's Address . . . . . . . . . . . . . . . . . . . . . . 7
Full Copyright Statement . . . . . . . . . . . . . . . . . . 8
1. Goals
The two goals of the CNRP [3] URI [1] are to identify both a specific
common-name record at a specific server and to identify a possibly
dynamic query or entry point into the query process. Since CNRP
requires that the ID be a core query term, these two cases can be
generalized down to simply specifying a query that contains only the
ID of the item.
On first glance it would seem a simple enough exercise to
canonicalize the XML encoded query and then insert it into the query
portion of the URL. The problem here is that, due to the encoding
rules, any remotely complex query will quickly blow out the URI
length limitations. The suggested solution is to provide a
simplified query syntax that is a subset of what is available via the
XML.
2. Terminology
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
document are to be interpreted as described in RFC 2119 [4].
3. Syntax Rules
3.1 General Syntax
The CNRP URI comes in two forms. The first form is for talking to a
specific server. The second form is for expressing a query that is
meant to be sent to several different CNRP services. The following
two examples are for pedagogical purposes only. The complete ABNF
grammar in Section 3.2 is the only authoritative syntax definition.
go://[<host>]?[<common-name>]*[;<attribute>=[<type>,]<value>]
and
go:<common-name>*[;<attribute>=[<type>,]<value>]
3.2 ABNF Grammar
The full ABNF [2] (certain values are included by reference from RFC
2396 [1]):
cnrp-uri = "go:" (form1 / form2)
form1 = "//" [server] ["?" ((common-name *avpair) / id-req) ]
form2 = common-name *avpair
id-req = "id=" value
avpair = ";" attribute "=" [ type "," ] value
server = // as specified in RFC2396
common-name = *(unreserved | escaped)
attribute = *(unreserved | escaped)
value = *(unreserved | escaped)
type = *(unreserved | escaped)
unreserved = // as specified in RFC2396
escaped = "%" hex hex
hex = "0" | "1" | "2" | "3" | "4" | "5" | "6" | "7" |
"8" | "9" | "A" | "B" | "C" | "D" | "E" | "F" |
"a" | "b" | "c" | "d" | "e" | "f"
3.3 Special Cases and Default Values
3.3.1 If There is Only a Server
In the case where the CNRP URI contains only the server production
then the URI identifies a given CNRP server, not any particular query
that is to be done. A client can assume that this server will at
least answer the 'servicequery' request.
3.3.2 If Server is Empty Then server=localhost
If the 'server' element has no value then its value MUST be assumed
to be "localhost".
3.3.3 Default Port
CNRP's well known HTTP transport port is 1096. If the port value
portion of the server production is not specified then port 1096
SHOULD be used if the client has no prior knowledge about other ports
or transports that the service may support.
3.4 Encoding Rules
The common-name, query parameters, and parameter values must be
encoded using the UTF-8 encoding scheme [5], and any octet that is
not one of the permitted characters per the above grammar MUST
instead be represented by a "%" followed by two characters from the
<hex> character set above. The two characters give the hexadecimal
representation of that octet.
4. Transport Independence
As stated in the CNRP protocol specification [3], CNRP is allowed to
be expressed over multiple transport protocols with HTTP being
mandatory to implement. In the case where a client attempts to
resolve a CNRP URI and it knows nothing about the service being
referenced in that URI, then it SHOULD use HTTP on the CNRP default
port (1096).
5. Examples
go:Mercedes%20Benz
This example shows a general query for the common-name "Mercedes
Benz". The intent is that the query should be packaged with any
client provided defaults and sent to the one or more services that
the client has configured to ask.
go://?Mercedes%20Benz
This example shows a general query for the common-name "Mercedes
Benz" that is sent to the server running on the 'localhost'.
go://cnrp.foo.com?Mercedes%20Benz;geography=US-ga
This example shows a query for the common-name "Mercedes Benz" in
the geographic area "US-ga" which should be sent to the server
found at cnrp.foo.com.
go://cnrp.foo.org?Martin%20J.%20D%C3%BCrst
This example includes a UTF-8 character encoded using hex
escaping. The value encoded is a u-umlaut (a 'u' with two dots
over it). This simple query is sent to a server found at
cnrp.foo.org with no parameters.
go://cnrp.foo.com?id=5432345
Here only an id is given which means that his example points
directly at a particular common-name record on a particular
server. This example would probably be found in a link on a web
page of some type.
6. Security Considerations
In addition to the security considerations inherent in CNRP itself
(see the Security Considerations section of RFC 3367 [3]), the URI
mechanism can also be used to retrieve a URI identifying some other
site by including just the ID and not the common-name being linked
to. I.e., the user may think he/she is being shown the URI currently
mapped to the "BMW" common-name but in the case where only the ID is
used the actual common-name is not part of the URI, thus making it
possible to use a CNRP URI without knowing which common-name it is
referring to.
7. IANA Considerations
The IANA is asked to register the URL registration template found in
Appendix A in accordance with RFC 2717 [6].
References
[1] Berners-Lee, T., Fielding, R. and L. Masinter, "Uniform Resource
Identifiers (URI): Generic Syntax", RFC 2396, August 1998.
[2] Crocker, D., "Augmented BNF for Syntax Specifications: ABNF",
RFC 2234, November 1997.
[3] Popp, N., Mealling, M. and M. Moseley, "Common Name Resolution
Protocol (CNRP)", RFC 3367, August 2002.
[4] Bradner, S., "Key words for use in RFCs to Indicate Requirement
Levels", BCP 14, RFC 2119, March 1997.
[5] The Unicode Consortium, "The Unicode Standard, Version 2.0:
Appendix A.2", ISBN 0-201-48345-9, January 1988.
[6] Petke, R. and I. King, "Registration Procedures for URL Scheme
Names", BCP 35, RFC 2717, November 1999.
Appendix A. Registration Template
URL scheme name: go
URL scheme syntax: Section 3.2
Character encoding considerations: Section 3.4
Intended usage: Section 1
Applications and/or protocols which use this scheme: [3]
Interoperability considerations: None not specified in [3]
Security considerations: Section 6
Relevant publications: [3]
Contact: CNRP Working Group
Author/Change Controller: IESG
Author's Address
Michael Mealling
VeriSign, Inc.
21345 Ridgetop Circle
Dulles, VA 20170
US
Phone: (703) 742-0400
EMail: michael@verisignlabs.com
Full Copyright Statement
Copyright (C) The Internet Society (2002). All Rights Reserved.
This document and translations of it may be copied and furnished to
others, and derivative works that comment on or otherwise explain it
or assist in its implementation may be prepared, copied, published
and distributed, in whole or in part, without restriction of any
kind, provided that the above copyright notice and this paragraph are
included on all such copies and derivative works. However, this
document itself may not be modified in any way, such as by removing
the copyright notice or references to the Internet Society or other
Internet organizations, except as needed for the purpose of
developing Internet standards in which case the procedures for
copyrights defined in the Internet Standards process must be
followed, or as required to translate it into languages other than
English.
The limited permissions granted above are perpetual and will not be
revoked by the Internet Society or its successors or assigns.
This document and the information contained herein is provided on an
"AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING
BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION
HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
Acknowledgement
Funding for the RFC Editor function is currently provided by the
Internet Society.
AI ELI5
On this page
§1Goals of the go: Scheme
The go: URI scheme is made for the Common Name Resolution Protocol (CNRP, RFC 3367). It has two goals: to identify a specific common-name record at a specific server, and to identify a possibly dynamic query or entry point into the query process.
Because CNRP requires the ID to be a core query term, both cases boil down to a query that holds only the ID of the item. The obvious idea, canonicalizing the XML query and inserting it into the URI query part, fails: with the URI encoding rules any remotely complex query would blow past URI length limits. So the RFC provides a simplified query syntax that is a subset of what the XML can say.
The abstract adds a warning: some components look like those in other URI schemes but do not act like them. The go scheme has more in common with the location-independent news scheme than any other.
Writing a whole question on a postcard takes too much room. So the go link is a short note that carries just the important parts of the question.
§3Syntax of a go: URI
A CNRP URI comes in two forms. Form 1 talks to a specific server: go://[server]?common-name;attribute=[type,]value. Form 2 expresses a query meant to be sent to several different CNRP services: go:common-name;attribute=[type,]value. These two patterns are only for teaching; the ABNF grammar in Section 3.2 is the only authoritative definition.
In the grammar, cnrp-uri is go: followed by form 1 or form 2. Form 1 is //, an optional server, and an optional ? part that holds either a common name followed by any number of attribute pairs, or an ID request written id=value. Each attribute pair is ;attribute=[type,]value. The server and the unreserved characters are taken from RFC 2396.
A go link has an optional address of who to ask, the name you are looking for, and optional extra clues like 'only in this place'.
§3.3Special Cases and Default Values
If the URI contains only the server part, it identifies a given CNRP server, not any particular query. A client can assume that such a server will at least answer the servicequery request.
If the server element is empty, its value MUST be assumed to be localhost.
CNRP's well-known HTTP transport port is 1096. If the port is not given in the server part, port 1096 SHOULD be used when the client has no prior knowledge of other ports or transports the service may support.
If you leave out who to ask, it means 'ask the computer right here'. If you leave out the door number, knock on door 1096.
§3.4Encoding Rules and Transport Independence
The common name, query parameters and parameter values must be encoded as UTF-8. Any octet that is not one of the characters the grammar permits MUST be written as % followed by two hexadecimal characters giving the value of that octet.
CNRP can be carried over multiple transports, with HTTP mandatory to implement. When a client tries to resolve a go: URI and knows nothing about the service it names, it SHOULD use HTTP on the CNRP default port, 1096.
Odd letters are written as a percent sign and two digits, like a secret code that any computer can read. If you do not know how a server likes to talk, use the standard way.
§5Examples
The RFC gives five examples. go:Mercedes%20Benz is a general query for a common name; the client packages it with its own defaults and sends it to the services it has configured. go://?Mercedes%20Benz sends the same query to the server on localhost.
go://cnrp.foo.com?Mercedes%20Benz;geography=US-ga adds a geographic attribute and targets the server at cnrp.foo.com. go://cnrp.foo.org?Martin%20J.%20D%C3%BCrst shows a UTF-8 character written with hex escaping. go://cnrp.foo.com?id=5432345 gives only an ID, so it points directly at one common-name record on one server, the sort of thing found in a link on a web page.
Examples show links that look up a car brand, one with a place filter, one with a letter that has two dots, and one that jumps straight to a numbered record.
go:Mercedes%20Benz
go://?Mercedes%20Benz
go://cnrp.foo.com?Mercedes%20Benz;geography=US-ga
go://cnrp.foo.org?Martin%20J.%20D%C3%BCrst
go://cnrp.foo.com?id=5432345§6Security Considerations
Beyond the security considerations of CNRP itself (see RFC 3367), the RFC points out a specific trap. A URI that holds only an ID, without the common name, can lead to a different site than the one a user expects.
A user may believe the link shows whatever is currently mapped to the common name "BMW", but since the name is not in the URI, a go: URI can be used without anyone knowing which common name it refers to.
If a link only has a secret number and no name, you cannot tell where it goes just by reading it.
§7IANA Registration
The IANA is asked to register the URL scheme using the template in Appendix A, following RFC 2717. The template names the scheme go, points to Section 3.2 for syntax and Section 3.4 for character encoding, and gives CNRP as the protocol that uses it.
It lists the CNRP Working Group as contact and the IESG as author and change controller.
The official list of link starting words needs to add 'go', so nobody else uses it for something different.