L2/12-178 Subject: Unicode Schedule From: Mark Davis Date: Mon, 30 Apr 2012 Unicode Member Representatives, The Unicode officers, in conjunction with the Unicode Editorial Committee, have recently been discussing the available options and schedule considerations for future releases of the Unicode Standard. We anticipate bringing this up for consideration at the upcoming May UTC meeting. In order to provide the Consortium members with a better opportunity for providing input on this issue in the context of their own anticipated requirements and potential product schedule constraints, we are circulating a summary of the options as we currently see them, together with our recommendation. The relevant considerations regarding anticipated character content for future versions are as follows. 1. The Turkish lira sign will likely be approved for encoding at the May UTC meeting, and is a high priority for early implementation. 2. Amendment 1 to 10646 3rd edition contains the repertoire of characters encoded for the Wingdings symbols, along with several non-controversial historic scripts and minority scripts. We will refer to this tranche as "Wingdings" for short, as those constitute the portion of Amendment 1 likely to be of most commercial relevance to product support decisions. 3. Amendment 2 to 10646 3rd edition currently contains most of the repertoire of characters encoded for the Webdings symbols, along with a larger number of historic scripts, including the controversial and highly visible Grantha script (for India) and the Old Hungarian script, along with several more minority scripts. We will refer to this tranche as "Webdings" for short. The relevant considerations regarding the ballot processing stages at which repertoire decisions are fixed and final, suitable to start processing for publication in a version of the Unicode Standard are as follows. Given current ISO rules, an amendment must be approved to start its FDAM ballot processing before we can count on no further technical changes occurring. 1. Amendment 1 will start its FDAM ballot following the October, 2012 WG2 meeting. The exact content is likely to be little different from what is currently in DAM ballot. 2. Amendment 2 will (most likely) start its FDAM ballot following the June, 2013 WG2 meeting. The content is much harder to predict right now, and may be substantially changed from what is currently in PDAM ballot. If changed, a likely outcome would simply be the removal from ballot of some controversial script or scripts. Taking those content and ballot schedule considerations into account, along with the anticipated upcoming schedule of UTC meetings, we can lay out three alternative scenarios for future Unicode Standard releases. Scenario A * Sept/Oct 2012 Unicode 6.2 (adding *only* the Turkish lira sign) * April 2013 Unicode 6.3 (adding the "Wingdings" tranche) * Jan 2014 Unicode 7.0 (adding the "Webdings" tranche) Scenario B * April 2013 Unicode 6.2 (adding the Turkish lira sign and the "Wingdings" tranche) * Jan 2014 Unicode 7.0 (adding the "Webdings" tranche) Scenario C * Sept/Oct 2012 Unicode 6.2 (adding *only* the Turkish lira sign) * April 2013 Unicode 6.2.1 (data fixes and limited UAX fixes only) * Jan 2014 Unicode 7.0 (adding both the "Wingdings" and "Webdings" tranches together) Note that the exact numbering for the Versions here ("6.2", "6.3", "7.0") is not fixed, and would ultimately be determined by how the UTC wants to position a particular release. The numbers are supplied here just for the sake of getting the comparison of scenarios formulated and into discussion. Scenario A would be the result of doing a hurry-up release for the Turkish lira sign, followed by more or less business-as-normal releases, doing a full release corresponding to the "Wingdings" tranche, and another full release corresponding to "Webdings" tranche. However, the assessment of the Unicode officers is that because of the schedules involved, the time between full releases (~5 months from 6.2 to 6.3, and then ~9 months from 6.3 to 7.0) is too tight for the available resources. It presents a very high risk of significant schedule slippage, and we do not recommend it. Scenario B would be the result of delaying the publication of the Turkish lira sign, and including it in the version for the "Wingdings" tranche. Our assessment is that this scenario would be feasible given the current resources, although the schedule between 6.2 and 7.0 under this scenario would be tight. The disadvantage of this scenario is that it delays the availability of the Turkish lira sign by 5 months, compared to Scenario A. Scenario C would be the result of doing a hurry-up release for the Turkish lira sign in Sept/Oct 2012, but then delaying further additions, so that the "Wingdings" and "Webdings" tranches would be accumulated into a large, combined repertoire addition for 7.0. In order to offset the long gap between the Sept/Oct 2012 Version 6.2 and the Jan 2014 Version 7.0, we would plan instead for a dot.dot release of Version 6.2.1 in April 2013. That dot.dot release would have no repertoire changes, but allow for rolling up any required property or other data file changes plus limited fixes for UAXes, if needed. Because the impact of a dot.dot release on editorial and other resources is more limited, it is our assessment that Scenario C is also feasible, with less risk of schedule slippage than Scenario A. The advantage of Scenario C is that it gets the Turkish lira sign published as soon as possible. It also releases both the "Wingdings" and the "Webdings" together, which may make more sense for both implementations and for marketing of the importance of the Unicode 7.0 release. Our recommendation is to proceed with Scenario C as the plan of record, unless there are countervailing requirements brought to the table by the Unicode members which may force a reevaluation of the feasibility and relative benefits of the other scenarios, or perhaps the consideration of other possible approaches.