BALLOT for "handling delimiting of patterns in complex messages"

[aphillips]

WORKING GROUP BALLOT

This issue being used to track votes on an issue as requested by the MessageFormat Subcommittee (working group) in the 2023-10-30 teleconference.

Please read the instructions CAREFULLY before responding.

The current syntax requires all patterns in complex messages to be delimited ("quoted") using curly brackets. The working group has discussed at length different options for allowing patterns to be unquoted. This document contains the proposed solutions, including some that have been rejected by the working group, along with some of the arguments related to the relative suitability of each solution.

Balloting Instructions

In the 2023-10-30 WG teleconference, there was a unanimous support for balloting the group on the following question:

Using instant runoff voting, rank your choice for how to handle delimiting of patterns in complex messages.

The deadline is 1700 (5 PM) in the America/Los_Angeles time zone on Saturday, 4 November, 2023

• A group member in good standing MAY submit one vote consisting of a ranked set of choices up until the deadline.
• A group member MAY edit, change, or delete their vote up until the deadline.
• Votes MUST be submitted as a comment on github issue #505. Group members who cannot submit a comment on this issue should contact the chair (@aphillips) for assitance.
• Votes MUST contain a stack ranked list of candidate options.
• Votes MUST only contain votes for candidate options. Write in votes are not acceptable.
• A vote MUST have at least one item in order to be counted and MAY rank two items or all three items.
• Only ranked votes will be counted. That is, do not submit a vote equating two entries.
• Group members MUST NOT comment on the votes of others in this issue. "Electioneering" or non-voting commentary is not permitted in the issue except the chair may seek clarification of a vote. You can, of course, discuss elsewhere. As a suggestion, please use #507 for this purpose)

Definitions

group member in good standing is any member of the MessageFormat mailing list or watcher of the message-format-wg github repo who has not be banned.

Candidates

Option 1 Always Quote Non-Simple Patterns (current syntax) Option 3 Permit non-simple patterns to be quoted and trim unquoted whitespace Option 4 Trim all unquoted whitespace, but do not permit quoting non-simple patterns

[aphillips]

(chair hat on)

This is an example of what a vote can look like.

(chair hat off and as a group member in good standing)

My stack rank preference is:

1 > 3 > 4

[duerst]

3 > 1 > 4

[sffc]

1 > 3 > 4

Prefer to keep the code mode syntax predictable. Option 3 has more than one way to do the same thing. Option 4 requires special syntax for leading and trailing whitespace which should "just work" without having to learn yet another corner of the syntax.

[sffc]

(moved to #507 by the chair)

[eemeli]

4 > 3

I wrote up some of my thoughts on this here: https://github.com/unicode-org/message-format-wg/pull/503#issuecomment-1784110857

[sffc]

I also don't understand this ballot. It is asking a narrow question (should we quote patterns in code mode) but the implications are much greater than that. Either option 2 or option 3 requires that we start quoting code in code mode. So we're not deciding a narrow question; we are deciding a direction from which future questions will need to be answered.

[gibson042]

4 > 3

[poulsbo]

1 > 3 > 4

[echeran]

FYI @aphillips created Issue #507 for a place to have discussion on this

[echeran]

1 > 3 > 4

[vdelau]

3 > 4 > 1

[mihnita]

1 > 3

[catamorphism]

1 > 3 > 4

[Crell]

3 > 1 > 4

[dminor]

3 > 4 > 1

[ryzokuken]

1 > 3 > 4

[SimonClark]

3 > 1 > 4

[relgu]

1 > 3 > 4

[flodolo]

4 > 3

[aurambaj]

1 > 3 > 4

[stasm]

1 > 3 > 4

[rxaviers]

1

[macchiati]

4 > 3 > 1

[aphillips]

===== voting closed here =====

I am using the IRV tool here: https://petertheone.github.io/IRV/

Results of this balloting will be reviewed in our teleconference call on Monday, 6 November, 2023.

[vdelau]

@aphillips I've ran the IRV tool but I'm not sure it is 100% correct. It does not seem to matter for the final result, but it doesn't seem to tally the first round correctly for me. I'm unable to attend today's call.

(For ballots, I filled out 2 vote ballots with the third option as least preferred, I removed the ballot with only a single vote.)

[aphillips]

@vdelau In running the tool, I filled out both 2-vote and 1-vote ballots with 0 for the unselected option(s). Not voting for a specific item or items is a real strategy in these sorts of multi-stage voting systems, when a voter does not wish to ever vote for a specific item or items. I therefore counted all 20 ballots and I did not infer any votes.

Because 11 of the 20 ballots have 1 first, we have an outright winner (regardless of voting system).

(Note: I also had a single transferable vote tool queued up, just speculatively, in case we ended in an IRV tie (which is, ironically, possible).)

[aphillips]

Group consensus is now #1. Thanks to all who participated in balloting.