[rfc-i] Removing postal information from RFCs [was: Changes to the v3 <postal> element]

Mark Nottingham mnot at mnot.net
Thu Jun 10 19:27:51 PDT 2021


> On 11 Jun 2021, at 12:21 pm, John Levine <johnl at taugh.com> wrote:
> 
> It appears that Mark Nottingham  <mnot at mnot.net> said:
>> Please clarify - do you intend to deprecate postalLine as well?
> 
> Yes, we're deprecating all of the <postal> subelements except <country>.

I don't see how deprecating <postalLine> follows from the problems you describe; if anything, they validate that it was the correct approach, and that the more detailed elements that were defined as an alternative were a mistake.

My understanding is that the rfc7991bis effort was focused on minimal changes; this doesn't seem aligned with that. What you're really doing here is proposing that postal information be removed from RFCs altogether, which is a pretty substantial change. I've adjusted the Subject line to better reflect this. 

Question: How many RFCs (text and XML) to date have postal information included? From what I've seen, it's a substantial portion. 

Cheers,


> 
> R's,
> John
> 
> 
>>> On 11 Jun 2021, at 9:54 am, John R Levine <johnl at taugh.com> wrote:
>>> 
>>> One of the changes to the xml v3 grammar since RFC 7991 is a new <postal> element with a set of subfields such as
>> <street>, <region>, and <code>. To render addresses we use a python library that depends on an open source address database
>> originally from Google.  While tracking down a rendering bug, we found that the rendering database is not actively
>> maintained and has a long list of unresolved pull requests.  We don't know of any other reliable source of rendering
>> patterns.
>>> 
>>> But we don't see a strong reason for readers to need the full postal address
>>> for RFC authors. Anecdotally, on rare occasions readers have used the postal
>>> address but (a) the email address is primary since <postal/> is optional and
>>> (b) readers likely have better ways to find contact information for RFC
>>> authors.
>>> 
>>> The one part we care most about from <postal/> is <country/>.  This enables
>>> people to do things like gather statistics about where RFCs originate, such as
>>> https://www.arkko.com/tools/rfcstats/d-countrydistr.html
>>> 
>>> Our proposal is to deprecate all of the <postal> elements other than the
>>> <country/> element.  Authors can include them if they want, but they won't be
>>> rendered and the RPC won't ask for them.  We think this leaves the useful bit
>>> of info while avoiding a lot of extra work for minimal benefit.
>>> 
>>> If there are issues we've missed, please let us know.
>>> 
>>> R's,
>>> John
>>> _______________________________________________
>>> rfc-interest mailing list
>>> rfc-interest at rfc-editor.org
>>> https://www.rfc-editor.org/mailman/listinfo/rfc-interest
>> 
>> --
>> Mark Nottingham   https://www.mnot.net/
>> 
>> _______________________________________________
>> rfc-interest mailing list
>> rfc-interest at rfc-editor.org
>> https://www.rfc-editor.org/mailman/listinfo/rfc-interest
> 
> 

--
Mark Nottingham   https://www.mnot.net/



More information about the rfc-interest mailing list