OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

xri message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [xri] Versioning ... was: Re: [xri] PLEASE REVIEW: Draft ofXRI-As-Relative-URI page


W3C Architecture of the World Wide Web.   It is the core document that the TAG references in there pronouncements.

http://www.w3.org/2001/tag/webarch/

It defines what they consider to be acceptable uses of the accept header etc.

=jbradley
On 15-Oct-08, at 10:17 AM, Barnhill, William [USA] wrote:

For the lurkers amongst us, AWWW = what again?

From: John Bradley
Sent: Wed 10/15/2008 12:47 PM
To: Nika Jones
Cc: OASIS XRI TC
Subject: Re: [xri] Versioning ... was: Re: [xri] PLEASE REVIEW: Draft of XRI-As-Relative-URI page

Using the accept header is something that we cant do because of AWWW.  We don't currently use the accept header for the proxy for anything other than service selection.

I am not opposed to having the proxy support multiple syntaxes for entering a XRI.

=jbradley
On 15-Oct-08, at 9:42 AM, Nika Jones wrote:

Both ideas make sense (headers/parameters) ... the only thing would be for people handwriting the HXRI (and would that acronym need to change for other schemes?) Most of the time I can see a client taking a native XRI and resolving it through the proxy (so the XRI to HXRI would be automated)... however there may be time when a person needs to write the HXRI by hand...

like OpenID... if I find that a client doesn't support native XRI (tisk, tisk *smile*) then I can type in the HXRI so that it conforms to the "http://*" that the OpenID client seems to be looking for, and that usually works. So for hand writing the XRI to an HXRI it seems easier to remember ( *.{TLD}/xri{version}:{XRI} )

example.net/xri2:=example or example.net/xri3:@example

than

example.net/=example?v=2

which is like ( *.{TLD}/{xri}{version-parameter} ) ... we'd also have very little HTTP Headers control when hand writing HXRIs. I don't want to dwell on it, but I just wanted to give some background as to where I was coming from.

Nika


On Oct 15, 2008, at 7:11 AM, John Bradley wrote:

For the http: proxy binding the XRI version should probably be in the query parameters.  That is where we specify returning XRD or XRDS etc.

Putting it in the base URI is going to cause problems.

=jbradley
On 15-Oct-08, at 6:32 AM, Barnhill, William [USA] wrote:


 

Just a couple of quick thoughts:

.. XRI resolution should be adaptable to non-http. I realize that's not currently supported, but should be IMHO. Please be careful that the changes you are making based on W3C feedback do not lock it totally to HTTP.

.. That said, if for the majority using HTTP you don't just have the URL to play with. The accepts header seems a good use for the version (the client specifies the versions it understands, the proxy returns the highest version the client understands). This also means URL links people have coded/made will not break when you version ('Cool URLs' and all that).

Thanks,

Bill



From: Nika Jones
Sent: Tue 10/14/2008 11:54 PM
To: OASIS XRI TC
Subject: Re: [xri] PLEASE REVIEW: Draft of XRI-As-Relative-URI page

With XRI 3.0 around the corner, would it be prudent to add a version to the sub-scheme?


or 


Nika


On Oct 14, 2008, at 6:09 PM, John Bradley wrote:

I think the idea is to move away from beginning the domain name with xri and move to making the significant part the tld.

For sftp we ned to sort out what the encoding rules are for the binding.  
Normally you would be adding query parameters to the URI as well so encoding for ftp may be a challenge.

=jbradley


On 14-Oct-08, at 6:00 PM, njones@ouno.com wrote:

All:

Does the XRI need to also be the first path... It's written as, "first
character of the path," so is it the first character of the first path?

would it be more appropriate to write it

*.{TLD}/{XRI}

where {TLD} is any Top Level Domain (.net, .com, .org, etc.) and the
{XRI}is any valid XRI?

Also for FTP (and FTPS and SFTP I presume) how does the resolution happen?
Would you need a special server listening to port 21 or 990? Or am I
getting confused as to where the FTP is placed. Is the following valid:

ftp://xri.net/@example*run

Nika


On Tue, 14 Oct 2008 00:05:31 -0700, "Drummond Reed"
<drummond.reed@cordance.net> wrote:
XRI TC Members and Observers:



Per my action item from the last telecon, I have posted a first draft
wiki
page describing the XRI-As-Relative-URI proposal to:



           http://wiki.oasis-open.org/xri/XriAsRelativeUri



The next step is for TC members to review and comment on this page. As
always, being a wiki, feel free to just go in and edit the page. Or, if
you
prefer, send feedback or question to the list.



The goal is to incorporate any feedback by Wednesday so we can begin
sending
messages to the W3C TAG public mailing list to solicit feedback there
before
our next telecon this coming Thursday.



Thanks,



=Drummond


---------------------------------------------------------------------
To unsubscribe from this mail list, you must leave the OASIS TC that
generates this mail.  Follow this link to all your TCs in OASIS at:
https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php







smime.p7s



[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]