suppress-script values for fil, mi, pes, prs, qu members

[Doug Ewell]

I appreciate Peter's quoting of relevant background material that led to 
the adoption of Suppress-Script, as opposed to Require-Script or no 
mechanism at all.

As clerk, I don't have a problem with preparing 48 sets of forms and 
records for these.

As an individual contributor, I'm concerned about the apparent precedent 
that would be set, in terms of trying to be comprehensive with S-S.

As Peter and others have noted, our knowledge is limited as to which 
languages are overwhelmingly written in one script, and which script 
that is.  Additionally, we have incomplete agreement among ourselves on 
the meaning of "overwhelming."  And the adoption of S-S was also 
influenced by resistance against the placement of the script subtag 
before region, because remove-from-right parsers built for RFC 1766/3066 
usage would not know to match "en-Latn-US" with "en-US".

So we've more or less settled on a conservative approach that takes into 
account whether the language tag would have been used in the RFC 
1766/3066 era—thus effectively excluding subtags introduced with ISO 
639-3—and whether the ietf-languages consensus on "overwhelming" is 
basically unanimous and unassailable.  For some of the S-S additions 
proposed by Peter, such as every language encompassed by Quechua, the 
second criterion is probably met, but I'm not sure about the first.

Currently we have 7844 language subtags, 134 of which have S-S.  Adding 
48 more (a 35% increase) does not pose any architectural problems, but 
could send us down a perfectionist rabbit hole where each and every 
language is subject to the S-S debate.  Ethnologue says that speakers of 
Bozaba ('bzo', population 5500) in the DRC "also use Lingala mainly in 
the market."  Since Lingala has an S-S of 'Latn', does it follow that 
Bozaba should as well?  Writers might be unlikely to use different 
scripts for the two.  Who wants to get involved in this debate?

If we want to try to be more comprehensive with S-S assignments, that's 
fine, but we'd better understand what we're getting into, and not be 
surprised when the requests start coming in one by one, or in flash 
floods.

--
Doug Ewell | Thornton, Colorado, USA | http://www.ewellic.org
RFC 5645, 4645, UTN #14 | ietf-languages @ is dot gd slash 2kf0s ­