- From: Greg Eck <greck@postone.net>
- Date: Sun, 10 Jan 2016 15:20:12 +0000
- To: "public-i18n-mongolian@w3.org" <public-i18n-mongolian@w3.org>
- Message-ID: <SN1PR10MB0943527D0DEE4A33E43A51B7AFC80@SN1PR10MB0943.namprd10.prod.outlook.com>
I am attaching the two documents that will go to the UTC. Please let me know of typos as well as suggestions. I have included the new U+1829-Isolate as it seems well attested to. I will be submitting these documents to Deborah Anderson in the next few days for her review also, so if there are areas not covered or typos, please let me know quickly. I have looked over the FVS Mis-Match area again. Badral, Jirimutu, I have tried to give it one more square and objective look as you have suggested. There are 7 areas of concern (1820, 1828, 182C(2x), 182D, 1835, 1836). But as we looked over the entire issue in our past discussions we found 10 more areas similar in nature (1822(2x), 1824(3x), 1826(3x), 185E, 1873). If we continue research in the same vein, we will find more. The DS01 (and other similar documentation in the future hopefully) documents the specification and the actuality of implementation, so it should be clear to new font developers how to do the implementation. There is a definite disconnect in the seven situations mentioned above. However, we are going to create more problems by trying to change the past documentation than leaving things as they are and explaining how the disconnect came about historically. All font developers agree on how the 7 areas are actually handled within the font/rendering_machine interaction. All of our fonts are working fine in these seven areas. I suggest that we just let it be and go on. Remember that the specification does not tell us where to implement the glyph context-wise - it only tells us where to place the FVS. The designers at the time of writing the MGWBM and the TR170 Report made a decision to consider everything between the initial letter and the final letter (MVS/NNBSP sequences included) as medials. Implementation time came around and the writers of Uniscribe, Harfbuzz and other rendering engines found it better to implement initial/final contexts around the MVS and NNBSP. Each font developer found a work-around the mis-match. We all understand the situation today. Each side has its own argument. Each side has a good argument. But the time has come to put the past behind us and move ahead. Let's not change anything more than we have to that has been set in the past. Greg
Attachments
- application/pdf attachment: DS01 Mongolian Base, Positional, Variant Forms 20160110.pdf
- application/vnd.openxmlformats-officedocument.wordprocessingml.document attachment: NP Proposed Changes.docx
Received on Sunday, 10 January 2016 15:20:49 UTC