[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [idn] iDNS transition: end-system vs. infrastructure?
- To: idn working group <idn@ops.ietf.org>
- Subject: Re: [idn] iDNS transition: end-system vs. infrastructure?
- From: Adonis El Fakih <adonis@ayna.com>
- Date: Sat, 12 May 2001 18:06:50 -0400 (EDT)
- Delivery-date: Sat, 12 May 2001 14:42:21 -0700
- Envelope-to: idn-data@psg.com
- User-Agent: Ayna VER 2.22
Hi,
I agree that there are various protocols that will need to be addressed case by case and examine the UTF8, UTF7, UCS2 or ACE solutions.
It would be easier if we can make a list of protocols/applications that will be affected by internationalized domain names (idn). I feel this would be the first step to make sure we did not skip any protocol/application in the process of evaluating what is the best solution.
>From the top of my head I can think of
HTTP requests -> DNS
Mail Clients -> SMTP or SMTP-8
Ftp CLients -> ASCII??
News Clients -> NNTP
IMAP CLients -> IMAP
POP CLients -> POP
Although we are looking at what technology to use be it plain UTF8, ACE or something in between, we also need to remember that the protocols we know today will slowly shift to support IDN, and in some cases some will never do.
I am sorry that I am trying to go to the basics, becuase both UTF8 and ACE solutions seem to be good depending on what apps we are talkign about, so if we define apps/protocols we are trying to make word with IDN, we may also have to introduce a time frame on using one or the other. i.e. it may be that ACE is the best way to implement something now, but then put recommendations forward that all new versions of each protocol should work with UTF8, becasuse ACE also has its drawbacks. In a way we are trying to map old systems to work with Unicode and future applications, and we want to force the additions of new value added data within the limited scope of 7bit ASCII to be fully backward compliant.
Sorry for the ranting. and Thanks.
Adonis
"Adam M. Costello" <amc@cs.berkeley.edu> كتب:
> Russ Rolfe <rrolfe@exchange.microsoft.com> wrote:
> could someone kindly post a simple list of the reasons they feel "Just
> do pure UTF-8" is *not* a valid alternative.
Domain names appearing in message headers, SMTP commands, URIs, etc. are
required by various standards to contain only ASCII letters, digits,
hyphens, and dots. Therefore there is no telling what existing software
would do if faced with domain names containing other characters. Some
of it would surely break. DNS is not the problem.
AMC
______________________________________________________________
Ayna, The Arabic Internet Starts Here. http://www.ayna.com