ietf
[Top] [All Lists]

Re: I-D ACTION:draft-brezak-spnego-http-00.txt

2001-09-25 12:30:02
On Tue, 25 Sep 2001, Garrett Wollman wrote:

This announcement continues a disturbing trend of MicrosoftSCII
appearing in what are supposed to be ASCII text documents.

I've observed this disturbing trend also.  Triggered by the formatting
problems and non-US-ASCII characters of a different draft, on Friday I
sent a message to the I-D administrator asking if she (and other
powers that be) would you be willing to institute a policy that draft
submissions that don't conform to the formatting guidelines would be
rejected.  My suggestion would be to provide a little program that
would look for these illegal characters and check the number of
characters per line and lines per page, and tell report whether the
submission passes or does not pass.  Alternatively, if the only
problem was the illegal quote and dash characters, it could simply
replace them.  [But one thing that puzzles me about these characters
is that sometimes they are in the range \221 - \227 where they appear
in the Windows character map and don't conflict with ISO-Latin-1, and
other times they are like \364 and \366 for double quotes which could
conflict with valid ô and ö characters in someone's name.]

She thanked me for my suggestion, but seemed not to be interested in
my offer to supply the format checking program.

I have a couple of questions regarding format policy:

  - Are unpaginated drafts acceptable?  I think they probably should
    be (and have been accepted in the past) because I-D's are supposed
    to serve for quick proposals as well as finished documents.  But
    if a draft does contain pagination, then it should be limited to
    58 lines as in the I-D guidelines and RFC 2223.

  - Is it accurate that I-D's and RFCs are supposed to contain only
    the US-ASCII (7-bit) character set?  Or are we supposed to allow
    for ISO-Latin-1 characters in people's names, etc.?  I thought I
    saw a document somewhere that answered this question (and said
    US-ASCII only), but RFC 2223 only says "The character codes are
    ASCII" which may not be specific enough.  The difficulty with
    allowing anything other than US-ASCII is that there is no place
    for a character set indication to go (this normally is placed in a
    MIME header, for example, and there is none on a plain text
    document).

I also noted that the I-D guidelines page itself does not conform to
the formatting requirements, with some lines much longer than 72
characters.  I provided a rejustified version, which might find its
way to the web site eventually.
                                                        -- Steve