Re: [xdi] Is it a problem that the XDI # symbol will conflict with URI # fragment char?

I've been thinking about this all week and I agree with Phil, Animesh, and Markus that, while we fought the good fight to try to keep XDI syntax entirely within the URL character set for 7+ years, it wasn't until we ultimately abandoned that constraint and expanded the character set that we could finally get some degree of readability back to the syntax.

At that point the decision was made: XDI addresses will need to be URL encoded when carried in URLs, and machines will need to do that URL encoding for humans.

Once we conceded that, then IMHO the primary criteria for selection for the XDI context symbols intended for ordinary humans to read and use (e.g., type into user interfaces) becomes:

How visually distinctive, attractive, and memorable are they?

Are there popular semantic associations already established with them, and if so, are they helpful or conflicting?

By these criteria, moving to + as the legal/business context symbol seems an easy choice. The only conflicts with + that we've found are with international phone numbers and potentially Google+. The former is minimal conflict because most +names will be entirely or primarily alpha characters. And the latter is proprietary and not really a namespace.

With # as the dictionary/tag context symbol, the visual appeal is clear, and the primary semantic association for most users will be hashtags, which the majority of the TC has decided is a helpful connotation due to Twitter and Facebook hashtags. The only conflict is with the fragment identifier in URLs, but that is a technical conflict and not a usability conflict (most users don't know what a fragment is).

So my recommendation is that we proceed with the Great Symbol Shift (so named by Andy Dale) of + for @, # for +, and @ for #.

I look forward to discussion and decision on tomorrow's call.

=Drummond 

On Wed, Mar 5, 2014 at 3:39 AM, Markus Sabadello <> wrote:

I agree with Animesh.. We already have lots of characters we need to encode. 

Markus

On Wed, Mar 5, 2014 at 2:04 AM, Chowdhury, Animesh <> wrote:

It is a concern. However, in the same token “+” , and “!” and “[“ and many other characters that can appear in an XDI address can’t be used in the URL with
 URL encoding either.

So, I’ve come to realize that an XDI address cannot be used in an URL without encoding. So, I don’t think making an exception for “#” is going to buy us much.

 

Thanks,

Animesh

 

From:  [mailto:]
On Behalf Of =Drummond Reed

Sent: Tuesday, March 04, 2014 7:19 PM

To: OASIS - XDI TC

Subject: [xdi] Is it a problem that the XDI # symbol will conflict with URI # fragment char?

 

As part of coming to a final decision about The Great Symbol Swap, I just had a long talk with Peter about his -1 vote. His primary concern is one I share: as attractive as # might be as the popular symbol for a hashtag—and thus generally
 applicable to the XDI dictionary space—it is also the URL fragment identifier.

 

This means that when it is typed in a URL, the browser does not send it to the server, but only processes it locally in parsing the returned resource.

 

So if we expect users to type #words (unescaped) as part of an XDI address into a browser address bar (without a plug-in), they will not be part of the XDI address sent to the server.

 

This is not a problem if we assume some form of XDI-aware smart client will be used to resolve XDI addresses (e.g., a browser plug-in, a web form submitted to an XDI resolver, a mobile app, etc.), as then all XDI addresses will be automatically
 URL-encoded.

 

So, I cast this question out to TC members: is this a problem you are concerned about or not?

 

=Drummond

 

On Tue, Mar 4, 2014 at 5:22 AM, Davis, Peter <> wrote:

-1.  

 

I remain concerned that we are just relocating the semantic collision from @ to # (both have meanings to the general population outside of XDI).

 

=peterd

 

On Mar 4, 2014, at 03:06 AM, Markus Sabadello <> wrote:

+1

 

On Tue, Mar 4, 2014 at 8:46 AM, =Drummond Reed <> wrote:

Per the minutes I just sent out, I have continued to check with both XDI TC members and others involved in related infrastructure with the following results:

There is unanimous support for reassigning + to the legal/business authority namespace (formerly represented by @).

There is nearly unanimous support for reassigning # to the dictionary namespace (formerly represented by +) and reassigning @ to the ordered collection member namespace (formerly represented by #).

So this is a call for rough consensus, i.e.:

If you agree with the proposed reassignments in the two bullet points above, reply with a +1.

If you are neutral, reply with a 0 (or don't reply).

If you disagree with these proposed reassignments, reply with a -1. 

Thanks,

 

=Drummond  

 

 

Peter Davis: Neustar, Inc.

Distinguished Engineer, Director, Neustar Foundry

45980 Center Oak Plaza Sterling, VA 20166

[T]
+1 571 434 5516 [E] 
 [W] 
http://www.neustar.biz/ [X] xri://@neustar*pdavis [X] xri://=peterd

The information contained in this e-mail message is intended only for the use of the recipient(s) named above and may contain confidential and/or privileged information. If you are not
 the intended recipient you have received this e-mail message in error and any review, dissemination, distribution, or copying of this message is strictly prohibited. If you have received this communication in error, please notify us immediately and delete
 the original message.