As I haven't received additional comments, I will send JGP's comments to IP in a separate mail. If you come up with further comments, please send them directly to IP. Hiro On Sun, 10 May 2015 00:01:27 +0900 HiroHOTTA <hotta@jprs.co.jp> wrote:
Dear CJK colleagues,
<regarding ToDo-14>
JGP drafted its comments to IP's feedback on Yoneya's algorithm. I hope CGP and KGP draft their own comments and send them to Hiro<hotta@jprs.co.jp> and Yoneya<yoshiro.yoneya@jprs.co.jp> by Monday night, 11 May. JGP will compile them all and send the comments to UP.
IP's feedback, which was sent to us at midnight of 28 April, is attached to this file for your reference.
Regards Hiro
======== (1) about the annotation on slide #3
Could you elaborate more on what this annotation means? Especially, we need to know exactly how 'script' should be used in RootLGR project.
For example, Japanese language uses Kanji, Hiragana and Katakana. In this case, is 'Japanese script' a proper usage? Or, is 'Kanji script' a proper usage?
(2) about WLE (to annotation on slide #4)
Does this annotation intend to be based on the situation "some of the WLE rules may be for strings that contain characters in multiple scripts"? For example, it refers to WLE rules for strings containing both Kanji and Hiragana, which are very common as Japanese words ?
The annotation says that "WLE rules are not script-specific in the procedure: there is only a single, common set.". Does this mean every LGR-2 has the same WLE? If so, development of RootLGR won't finish until the last Script LGR finished. Or does it mean every LGR-2 which shares the same script has the same WLE? If the answer is latter, CJK coordination steps has to have WLE integration step which will require human intervention. I would like to see the concrete image of LGR-2.
(3) about the term 'disposition' (to annotation on slide #6)
The annotation says that "Note: in the XML schema this is done via the "type" attribute, on the <var> element and only labels have dispositions". When we say "disposition" in LGR context, does it mean label disposition? How can we say "attribute on the <var> element" shortly? Attribute? Type? Disposition type? It is very helpful if we (IP and GPs) use the same terminology.
(4) about (o) (disposition) type (to annotate on slide #11)
The annotation says that "Should be (b)?". Someone feels it should be (o), and someone feels it should be (b). Whether it is (b) or (o) doesn't make the result of WLE different (the label is invalid). Is our understanding correct?
(5) about minimization of allocatable variant labels (to annotate on slide #12 & #13 & #16)
The annotation says that "Overall, the goal of the IP is to help the CJK-GPs to arrive at coordinated LGRs, such that the number of allocatable variant labels is minimized.". My understanding is that selection of delegated label(s) from "allocatable label(s)" is by decision of applicant. And it is out of scope of RootLGR.
For Japanese language, it is impossible to predict which character is allocatable and which is blocked individually because all characters are allocatable independently. It depends on the applicant's thought. JGP doesn't say all the variant labels should be delegated. The number of delegated labels will be at most 2 or 3. But, for example, which labels are appropriate among the 16 variant labels to be allocated cannot be predefined in the RootLGR level. Such limitation should be done in application proposal and its evaluation because all the variant labels are valid in Japanese case.
As a summary, we should not try to solve all the issuees by RootLGR, we should think about appropriate mixture of RootLGR and application evaluation rules/guidelines.
(6) about Han and Kana mixed labels (to annotate on slide #16)
The annotation says that "What about the case where a label is applied for that combines Han and Kana characters in Japanese. Which cases do we expect that the Han-variant can be substituted for the applied for Han code point in such a mixed label? -- This issue is specific to JGP, but it is affected by the design of the Han variants -- unless JGP has a WLE rule to "block" variants for mixed labels altogether.". For Japanese language, Han and Kana can use with any combination. As mentioned above, it is impossible to predict which character is allocatable and which is blocked individually. Mixed use of Han and Kana does not affect it.
========
On Fri, 08 May 2015 17:11:04 +0900 HiroHOTTA <hotta@jprs.co.jp> wrote:
Dear CJK Colleagues,
Below I marked the progress status to each ToDo item except ToDo-1 to 4, which request each participant to build his/her own capacity about RootLGR. I put the progress status after "==>" following each ToDo in the list pasted below..
Please check the progress status of your ToDo's. If correction is needed, please let me know.
I added one ToDo which is == ToDo-14<expected> each of CGP/JGP/KGP reviews the feedback from IP (a result of ToDo-5), drafts questions and comments to IP, and sends it to Hiro<hotta@jprs.co.jp> and Yoneya<yoshiro.yoneya@jprs.co.jp> by Monday night, 11 May and then, Hiro and Yoneya will compile the questions and comments into one file and send it to IP on Tuesday morning
I added this because I believe we need to raise questions and comments to IP's response given in ToDo-5. I attach the subject document, which is the IP's response, for your reference.
I hope you will work on this ToDo-14 and send your questions and comments to Hiro<hotta@jprs.co.jp> and Yoneya<yoshiro.yoneya@jprs.co.jp> by Monday night, 11 May.
Regards, Hiro
========================================================================
ToDo's 2nd draft (as of noon KST on 8 May)
==
[[Premise]]
ToDo-1 <must> Each participant understands what RootLGR is and what is expected for GPs to do.
ToDo-2 <must> Each participant understands Yoneya's algorithm that was already sent to CGP/JGP/KGP by Yoneya and also agreed by C and J in Dallas, which is attached to this mail as well
ToDo-3 <must> Each participant understands MSS concept that was already sent to CGP/JGP/KGP by Dr. Wang Wei, which is attached to this mail along with HiroHOTTA's response
ToDo-4 <expected> Participants agree on Yoneya's algorithm as a framework and also agree on partial usage of MSS to accelerate our discussion ("partial" means "J doesn't need to be considered to be incorporated into MSS") This is expected to be discussed and finalized online before our meeting
[[Integration Algorithm]]
ToDo-5 <expected> J gets IP's feedback on Yoneya's algorithm
==> done
[[MSS/LGR-1]]
ToDo-6 <must> C prepares MSS repertoire, which may be equivalent to Chinese LGR-1 repertoire (done?)
==> done
ToDo-7 <expected> C prepares Chinese variants within MSS, which may be equivalent to Chinese LGR-1 (planned date is expected to be declared, if not in time for the meeting)
==> done
ToDo-8 <must> J prepares Japanese LGR-1 repertoire and variants (there's no variants in Japanese LGR-1 : they were already sent to CGP/JGP/KGP)
==> done
ToDo-9 <must> K prepares the basic idea of Korean LGR-1 repertoire and variants
==> working (hopefully to be done by 14 May)
ToDo-10<expected> K prepares Korean LGR-1 repertoire and variants (planned date is expected to be declared, ift LGR-1 does not come in time for the meeting)
==> working (hopefully to be done by 14 May)
ToDo-11<expected> each of CGP/JGP/KGP assesses the repertoires and variants that have already been provided by other GPs as far as possible
==> working (C and J are assesing each other)
[[Logistics/etc.]]
ToDo-12<must> each CGP/JGP/KGP Chair designates a person in charge of ToDo-5 to ToDo-11 well in advance to the meeting (expected to post the (names) in replying this mail by May 1st) - this may accelerate the coordination a lot
==> working (C:Wang Wei, J:Yoneya, K:???)
ToDo-13<must> convener fixes the agenda through consultation with CJK colleagues (Hiro is pleased to behave as the convener until someone will raise his/her hand)
==> working (to be fixed Monday night, 11 May)
[[Integration Algorithm - additional]]
ToDo-14<expected> each of CGP/JGP/KGP reviews the feedback from IP as a result of ToDo-5, and drafts questions and comments to IP and sends it to Hiro<hotta@jprs.co.jp> and Yoneya<yoshiro.yoneya@jprs.co.jp> by Monday night, 11 May, and then Hiro and Yoneya will compile the questions and comments into one file and send it to IP on Tuesday morning
==> just begun
========================================================================
On Sun, 26 Apr 2015 02:05:17 +0900 HiroHOTTA <hotta@jprs.co.jp> wrote:
Dear colleagues in CGP/JGP/KGP,
If I may, in order for us to make our Seoul meeting efficient and fruitful, I'd like to propose what each of us is expected to prepare well before the meeting.
I know I am very pushy but I think at least we must not use our precious time just to understand the information in front of us for a long time.
Please give comments and let's discuss online about the ToDo's before Seoul meeting .
==
[[Premise]]
ToDo-1 <must> Each participant understands what RootLGR is and what is expected for GPs to do.
ToDo-2 <must> Each participant understands Yoneya's algorithm that was already sent to CGP/JGP/KGP by Yoneya and also agreed by C and J in Dallas, which is attached to this mail as well
ToDo-3 <must> Each participant understands MSS concept that was already sent to CGP/JGP/KGP by Dr. Wang Wei, which is attached to this mail along with HiroHOTTA's response
ToDo-4 <expected> Participants agree on Yoneya's algorithm as a framework and also agree on partial usage of MSS to accelerate our discussion ("partial" means "J doesn't need to be considered to be incorporated into MSS") This is expected to be discussed and finalized online before our meeting
[[Integration Algorithm]]
ToDo-5 <expected> J gets IP's feedback on Yoneya's algorithm
[[MSS/LGR-1]]
ToDo-6 <must> C prepares MSS repertoire, which may be equivalent to Chinese LGR-1 repertoire (done?)
ToDo-7 <expected> C prepares Chinese variants within MSS, which may be equivalent to Chinese LGR-1 (planned date is expected to be declared, if not in time for the meeting)
ToDo-8 <must> J prepares Japanese LGR-1 repertoire and variants (there's no variants in Japanese LGR-1 : they were already sent to CGP/JGP/KGP)
ToDo-9 <must> K prepares the basic idea of Korean LGR-1 repertoire and variants
ToDo-10<expected> K prepares Korean LGR-1 repertoire and variants (planned date is expected to be declared, ift LGR-1 does not come in time for the meeting)
ToDo-11<expected> each of CGP/JGP/KGP assesses the repertoires and variants that have already been provided by other GPs as far as possible
[[Logistics/etc.]]
ToDo-12<must> each CGP/JGP/KGP Chair designates a person in charge of ToDo-5 to ToDo-11 well in advance to the meeting (expected to post the (names) in replying this mail by May 1st) - this may accelerate the coordination a lot
ToDo-13<must> convener fixes the agenda through consultation with CJK colleagues (Hiro is pleased to behave as the convener until someone will raise his/her hand)
Hiro
_______________________________________________ japanesegp mailing list japanesegp@icann.org https://mm.icann.org/mailman/listinfo/japanesegp
_______________________________________________ japanesegp mailing list japanesegp@icann.org https://mm.icann.org/mailman/listinfo/japanesegp