[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: RE: [ciq] follow up on David Putman's comments
"COMMENT"s in body of you Email. -----Original Message----- From: Max Voskob [mailto:max.voskob@paradise.net.nz] Sent: Monday, May 02, 2005 5:19 AM To: ciq@lists.oasis-open.org Subject: [ciq] follow up on David Putman's comments Hi David, I was going thru your comments that I printed out some time ago. I think I reflected some of your suggestions, but I'd like to clarify some others. --- Order of elements --- On the one hand the order of free text address must be preserved, on the other, the presentation is locale dependent. I suggest the only requirement: 1. The order must be preserved How the address is presented is for the implementation to decide. Ultimately, it is the humans that will use the address or the address must go thru a parser to be used by machines for some sorting or delivery purposes. COMMENT: Okay, but that order SHOULD preserve all of the data within the address. That SHOULD include punctuation (some of which is critical to the "meaning" of the address!). ------------------------- --- Administrative area vs. Locality --- I agree that the boundary is very vague. Shall we merge them and make a hierarchical structure where one area is included in the other? It is unlikely to go down further than 3 levels. COMMENT: Unfortunately, "postal standards" as evidenced in some UPU examples and many other "common use" examples do include more than three (at least for "administrative area" and perhaps for "locality". The only way to bypass this is to either create an unbounded structure for their instantiation or just lodge them in undifferentiated (not split up) address lines - even in the detailed model! ---------------------------------------- --- Postal elements --- What are the alternatives to the current structure? You are saying it's not that good, but I can't think of anything better. Any suggestions? COMMENT: Yes, I am suggesting that there be a specific sub-structure and/or types / attributes to identify those - as those are SOMETIMES the complete and only representation of an address. This "complete and only" is mostly contained in the address lines NOT associated with city, state, and country ("adminstrativer areas"). However, "unique ZIP" type addresses in the USA can be comprised only and totally of what one might term "postal elements". That is, only a ZIP code (a postal element) is required and the "city" can actually be a firm name. In USA internal addressing (as with most other countries!), NO country is present (only implied). ----------------------- --- Care of --- Look, I'm confused myself here who's in care of who. We need revisit this with more use cases. --------------- COMMENT: Okay. But basically, ANYONE / ANYTHING can be put in (after) a "care of" indication for any one of a number of valid and needed purposes. It might be the person normally residing at an address (say, accepting delivery for a guest who is designated as the primary recipient). It can be another person / function / department at a business address for any one of a number of reasons (defunct department; person acting for a person on leave; temporary running of a function in another area, etc.); it can also be used for specifying a physical or logical "mail stop" to which the mail should be delivered (a way of routing to the intended recipient via company-internal, interoffice mail). It might be a services organization for a person (private mail box companies) or business (bill receiving agency; so that "care of" specifies the "real recipient" (by name, code / account number, etc.), while the service provider recipient actually receives the mail and holds or forwards it to their customer (the address provided IS "properly" associated with the service provider recipient!). For General Delivery addresses, it can specify WHO can pick up the mail at a post office other than the primary recipient. --- Examples --- Most of the examples were made up leading to poor diversity. What we need is a bunch of realistic and diverse examples. Let's create and maintain a doco with examples. It will allow us to rebuild test cases when the schema changes. ---------------- COMMENT: Good and agreed. Cheers, Max --------------------------------------------------------------------- To unsubscribe from this mail list, you must leave the OASIS TC that generates this mail. You may a link to this group and all your TCs in OASIS at: https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]