Jump to content

Talk:URI scheme: Difference between revisions

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Content deleted Content added
→‎Split article: new section
Line 120: Line 120:


The article misses to describe [[Percent-encoding]] and the allowed character set. What characters is an URI allowed to be build of at which parts? -- [[Special:Contributions/195.37.139.208|195.37.139.208]] ([[User talk:195.37.139.208|talk]]) 10:18, 18 June 2009 (UTC)
The article misses to describe [[Percent-encoding]] and the allowed character set. What characters is an URI allowed to be build of at which parts? -- [[Special:Contributions/195.37.139.208|195.37.139.208]] ([[User talk:195.37.139.208|talk]]) 10:18, 18 June 2009 (UTC)

== Split article ==

I suggest splitting this article into two new articles:
1) Explaining URI scheme in general
2) List of URI schemes

Revision as of 22:01, 7 October 2009

This is the talk page for discussing changes to the URI scheme article.

  • Please do not use it as a forum for general discussion about the article's subject.
  • Sign and date your posts using four tildes (~~~~).
  • Place new comments after existing ones (but within topic sections).
  • Separate topic sections with a ==Descriptive header==.

Double slashes

The double slash after the colon following the URI scheme name requires some explanation IMHO. Why is it used sometimes (e. g. in http) but not always (e. g. in mailto or news)? Why do Windows "file:" URIs for UNC paths have to use four slashes, like in "file:////myserver/myshare/myfile.htm"; couldn't they live with the UNC's two slashes as well? Also, the scheme list could need some examples. - wr 14-dec-2005

For Unixy paths, I believe it is e.g. "file:///home/isaac/whatever" because it's "file://" + "/home/isaac/whatever". I don't know why they have the "//", but a bit of searching turns up this: "The scheme specific data start with a double slash "//" to indicate that it complies with the common Internet scheme syntax." (RFC 1738) —Isaac Dupree(talk) 23:31, 1 January 2006 (UTC)[reply]
It seems like the revised standard in RFC 3986 takes a rather different approach than its predecessors to the generic-ness of URIs. In the previous versions, very little meaning was explicitly enforced on the structure of the "scheme-specific part", and those schemes which used the common hierarchical system were referred to as using a "generic URI" syntax. Part of this generic syntax was the leading "//", so a URI starting <scheme>:// (e.g. http://) could be assumed to be using that syntax, while one that didn't was probably non-hierarchical (e.g. mailto:). Excerpt from RFC 2396:
   The URI syntax does not require that the scheme-specific-part have
   any general structure or set of semantics which is common among all
   URI.  However, a subset of URI do share a common syntax for
   representing hierarchical relationships within the namespace.  This
   "generic URI" syntax consists of a sequence of four main components:

      <scheme>://<authority><path>?<query>

   each of which, except <scheme>, may be absent from a particular URI.
   For example, some URI schemes do not allow an <authority> component,
   and others do not use a <query> component.
How this all fits in with the different angle taken by the newer RFC, I've no idea; I'd have to read it first.
As for file://, Isaac's given most of the answer already: a UNIX path becomes file:///path/file from file:// + /path/file - that is, the file: scheme with the marker that it's generic/hierarchical (//), and then the path used on the local system. Similarly, a Windows UNC example like file:////myserver/myshare/myfile.htm is simply file:// + //myserver/myshare/myfile.htm. - IMSoP 13:41, 29 August 2006 (UTC)[reply]
perhaps it was to make it easier for parsers to differentiate between an authority and a relative path, when certain elements are missing. For instance, //example.com/ clearly refers to an authority, while /example.html/ clearly is a path. -Cowlinator —Preceding comment was added at 02:27, 27 October 2007 (UTC)[reply]

steam:// URI

I could add it meself, but I'm lazy...Takua108 02:15, 26 April 2006 (UTC)[reply]

Added as of 12 September 2006 -- Southen 17:49, 20 October 2006 (UTC)[reply]

require cleanup??

"To meet Wikipedia's quality standards, this article or section may require cleanup".

Clean the Notes column is enough??

I think we nedd information, other articles like URI does not offer information about schemas. Other sugestion is to split into a lot of schema articles. -- krauss 31/jul/2006.

Splitting the Table

The List of URI schemes table is getting long and complex. Is there a rule that prevents us from splitting it into two sections, for "Official IANA-registered schemes" and "Unofficial but common schemes"?

This would have the benefit of having both tables appearing in the table of contents for easy access, as well as make it easier to edit the required table. Thoughts? -- Techtoucian 04:39, 19 October 2006 (UTC)[reply]

I support splitting the IANA and unofficial schemes into two tables (that is, each with their own heading) on this same page. The unofficial schemes could probably do with alphanumeric sorting too. -- Southen 17:49, 20 October 2006 (UTC)[reply]
Out of interest, is there any particular reason why no-one thought to sort the table of registered schemes while they were at it? Oh well, I'm going to do it now, anyway. - IMSoP 17:46, 6 January 2007 (UTC)[reply]

info-URI

What about info: (RFC 4452)

Official IANA-registered schemes

I have some suggestions about Official IANA-registered schemes, but I would like to discuss them because they may imply substantial modifications:

  • I think the "Notes" column is too narrow (because "General format" needs most of the width), may be there should not be a table but a list ...
  • Purpose and Name are not filled with uniform criterium on the whole table (usually "purpose" is the "Name" of the protocol, or the "Meaning" of the schema identifier, but there are some exceptions). On some entries the "Notes" field is really the "Purpose" of that schema.
  • How should the general format be stated? ABNF per specification? "simplified" ABNF? example or meta-example (such as dns://dnsauthority/dnsname?dnsquery).
  • "generic syntax" is ambiguous, and should be avoided when possible (e.g. http:, ftp:, etc).

Rjgodoy 01:24, 4 April 2007 (UTC)[reply]

Use Backus–Naur formatting

Authors are encurrage to use Backus–Naur formatting when describing the URI format. A generic example would be:

<example_uri> ::= "example://" <absolute_uri>

— Preceding unsigned comment added by 80.202.179.68 (talk)

Good point (and well done too!). Rjgodoy 01:44, 7 August 2007 (UTC)[reply]


icyx://, rtpx://, htpx://, uvxx://

Would it be appropriate to add these unofficial URLs: icyx://, rtpx://, htpx://, uvxx:// or aren't they yet common enough? They are used by Orban/Coding Technologies AAC/aacPlus plugin for any DirectShow compliant media player, such as Windows Media Player.

More information.

24.80.185.126 (talk) 12:06, 3 September 2008 (UTC)[reply]

Picasa?

Doesn't Picasa use picasa: as a URI scheme? It isn't listed in the "unofficial" table, or in the template at the bottom of the article. B7T (talk) 15:28, 17 March 2009 (UTC)[reply]

URI not listed yet

1. reg syntax : reg:[regpath]

    ex : reg:HKEY_CURRENT_USER\Control Panel OR reg:HKCU\Control Panel

Logiphile (talk) 15:04, 22 March 2009 (UTC)[reply]

res

Outlook seems to embed images as res://ietag.dll/#34/#1001 - but that scheme isn't listed on this page? 80.177.58.134 (talk) 09:20, 5 May 2009 (UTC)[reply]

Percent-encoding and character set

The article misses to describe Percent-encoding and the allowed character set. What characters is an URI allowed to be build of at which parts? -- 195.37.139.208 (talk) 10:18, 18 June 2009 (UTC)[reply]

Split article

I suggest splitting this article into two new articles: 1) Explaining URI scheme in general 2) List of URI schemes