RE: [lemonade] POLL: discovery of available conversions in CONVERT

[<[e-mail]>]

Hi Alexey,
please find my answers below.

Q1:
===
A - yes (in case mailbox annotation is supported)
B - yes (always)
C - no (if there is none though, unicode-casemap is mandatory)
D - yes (though not explicitly stated -> improve text?)

Q2:
===
2. Note: it looks good on paper, the question is how efficient and useful this will be in practice (when will a client declare it, client has to remember that it has declared it already, what happens if it was not declared, etc).

3. While Dave's proposal is good, it might seem a logical choice to store the client capabilites as METADATA (or, some other form) so that it does not need to be disclosed during each session. This seems like a good candidate for the client identification draft as well. Some vendors expressed their concerns about the METADATA extension though - so it might not be such a good idea in the end.

Q3:
===
1. If "I do not know what this thing can be turned into", it might be more efficient to ask the server "just give me something I understand".

5. I suggested earlier to allow the client to request anything, and return the possible conversions only when there was an error. Arnt's proposal (advertise the list of MIME types) would imrove this: the MIME type chosen by the client will be always correct and only the parameters could be wrong.

Q4:
===
4. See improvement suggestion (using Arnt's proposal) above: Q3/5.

Q5:
===
I think that answer Q4/4 implies that it would be mandatory.

Best regards: Zoltán Ördögh
E-mail: zoltan dot ordogh at nokia dot com
Phone: [phone]

 

>-----Original Message-----
>From: ext Alexey Melnikov [mailto:[e-mail]] 
>Sent: 25 January, 2008 20:03
>To: Lemonade WG
>Subject: [lemonade] POLL: discovery of available conversions in CONVERT
>
>Hi everyone,
>After reviewing comments on CONVERT capability discovery 
>received during and after WG LC it looks like there is no 
>consensus (at least not yet) to use any particular mechanism. 
>Please correct me if you think otherwise.
>
>So in order to figure out if the WG prefers any particular 
>mechanism I suggest that people answer the following 
>questionnaire. Some questions allow for multiple choices. 
>Please list them in the order of your preferences.
>
>In order to correctly answer the questionnaire, one needs to:
>1). have basic understand about CONVERT, ideally read 
>draft-ietf-lemonade-convert-13.txt
>2). review draft-daboo-imap-annotatemore-12.txt
>3). review draft-melnikov-lemonade-convert-discovery-00.txt
>
>Please either send your replies to the Lemonade WG mailing 
>list, or if you prefer - directly to me, CCing Peter Coates 
><[e-mail]>, Eric Burger <[e-mail]>, Glenn 
>Parsons <[e-mail]>.
>
>Please send your response before February 8th.
>==========================================
>
>Q1: Some questions regarding understanding of METADATA-SERVER
>(draft-daboo-imap-annotatemore-12.txt) extension (please 
>answer to each question with Yes (true) or No (false))
>
>A). METADATA-SERVER depends on LIST-EXTENDED extension 
>(draft-ietf-imapext-list-extensions-18.txt).
>B). METADATA-SERVER depends on ENABLE extension 
>(draft-gulbrandsen-imap-enable-05.txt).
>C). A server implementing METADATA-SERVER must implement 
>multiple comparators.
>D). METADATA-SERVER requires implementation of both .priv and .shared
>
>
>Q2: Should CONVERT support the following conversion discovery 
>mechanism: 
>"Give me this body part in a format that I probably understand"
>(feel free to select multiple answers, unless you answer No or 
>Undecided)
>
>Answers:
>1). No
>2). Yes, I am happy with Dave Cridland's proposal (see section 
>2.1 of draft-melnikov-lemonade-convert-discovery-00.txt or
><http://www1.ietf.org/mail-archive/web/lemonade/current/msg04264.html>)
>3). Yes, but I would like to suggest another proposal 4). Undecided
>
>
>Q3: Should CONVERT support the following conversion discovery 
>mechanism: 
>"What can I turn this body part into?"
>(feel free to select multiple answers, unless you answer No or 
>Undecided)
>
>Answers:
>1). No
>2). Yes, I prefer to use METADATA-SERVER extension (see 
>section 2.2.1 of 
>draft-melnikov-lemonade-convert-discovery-00.txt, also see 
>draft-daboo-imap-annotatemore-12.txt for more information) 3). 
>Yes, I prefer to use Dan Karp's proposal (see section 2.2.2 of 
>draft-melnikov-lemonade-convert-discovery-00.txt or
><http://www1.ietf.org/mail-archive/web/lemonade/current/msg04297.html>)
>4). Yes, I prefer Arnt's proposal for advertising target MIME 
>types (or similar tokens) in the CAPABILITY response, e.g.:
>
>* CAPABILITY ... CONVERT=JPEG ... CONVERT=MP3 ...
>
>5). Yes, but I would like to suggest another proposal 6). Undecided
>
>
>Q4: Should CONVERT support discovery of available conversion 
>parameters for a particular conversion?
>(feel free to select multiple answers, unless you answer No or
>Undecided) Note that a separate question is asking if 
>discovery should be optional or mandatory for servers to implement.
>
>Answers:
>1). No, reporting that a particular conversion parameter is 
>not supported (see CONVERT draft) is sufficient 2). Yes, I 
>prefer to use METADATA-SERVER extension (see section 2.2.1 of 
>draft-melnikov-lemonade-convert-discovery-00.txt, also see 
>draft-daboo-imap-annotatemore-12.txt for more information) 3). 
>Yes, I prefer to use Dan Karp's proposal (see section 2.2.2 of 
>draft-melnikov-lemonade-convert-discovery-00.txt or
><http://www1.ietf.org/mail-archive/web/lemonade/current/msg04297.html>)
>4). Yes, I would prefer to list available conversion 
>parameters upon a CONVERT command failure caused by an 
>unrecognized conversion parameter, for example:
>
>      C: b002 CONVERT 2 ("text/plain" ("mumble" "hrrr"))
>          (BINARY[3] BODYPARTSTRUCTURE[3])
>      S: * 2 CONVERTED (tag "b002") (BODYPARTSTRUCTURE[3]
>          (ERROR "Parameter mumble is not supported for text/plain" 
>BADPARAMETERS
>          "text/html" "text/plain" ("mumble" "hrrr")))
>      S: * CONVERSIONS "text/html" "text/plain" "charset" 
>"UNKNOWN-CHARACTER-REPLACEMENT"
>      S: b002 OK Some conversions failed  
>
>5). Yes, but I would like to suggest another proposal 6). Undecided
>
>
>Q5: If you answered Yes to Q4: should the conversion parameter 
>discovery be Mandatory or Optional for servers to support?
>
>
>
>_______________________________________________
>lemonade mailing list
>[e-mail]
>https://www1.ietf.org/mailman/listinfo/lemonade
>Supplemental Web Site:
>http://www.standardstrack.com/ietf/lemonade
>


_______________________________________________
lemonade mailing list
[e-mail]
https://www1.ietf.org/mailman/listinfo/lemonade
Supplemental Web Site:
http://www.standardstrack.com/ietf/lemonade