At Thu, 07 Feb 2008 03:06:43 -0500,
John C Klensin wrote:
--On Thursday, 07 February, 2008 09:28 +0200
Pasi(_dot_)Eronen(_at_)nokia(_dot_)com wrote:
John C Klensin wrote:
For the second, it claims that the file isn't "plain text" and
won't post it or even provide a manual submission path. The
file is output or xml2rfc, has proper CRLF line endings, and,
if it contains any non-ASCII characters or serious format
misbehavior, the online version of idnits can't find it, even
in very verbose mode. [rt.amsl.com #1799]
I also encountered this problem (rt.amsl.com #1730), and after
some debugging, discovered what the problem was: the
submission tool uses the "file" program to check whether the
file is plain text or not, and (apparently) the "file" program
on the new servers behaves slightly differently from the old
one.
In particular, if you have the string "(if approved)" on your
cover page, some versions of "file" (at least some Linux
distributions) will identify your draft as "Lisp/Scheme
program text" instead of just ASCII :-)
<obscenity/>
So we get a choice of either
* Not getting drafts posted or
* Violating the IESG's instructions about how I-Ds that
are intended for standards track are laid out.
Gee.
Ray, while meeting site choices are ultimately judgment calls
and about tradeoffs, this is unambiguously preventing work from
getting done. Today is Thursday and we've got a posting cutoff
a week from Monday.
It's also worth noting that manually posted drafts (which is
what I fell back to after running into this bug) do not appear
to be showing up especially fast. draft-ietf-tls-rfc4346bis-09
was submitted after COB on Mon and isn't in the repository
yet.
Obviously, it's not a disaster if drafts which are postd
through exceptional paths take 2+ days to show up, but if
that's going to be the normal path, that's a little less
attractive.
-Ekr
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
http://www.ietf.org/mailman/listinfo/ietf