Re: [Tools-discuss] Re: RFCmarkup v1.28

[Henrik Levkowetz <henrik at levkowetz.com>]

Hi Frank,

on 2006-07-28 13:21 Frank Ellermann said the following:
> Henrik Levkowetz wrote:
> 
>> please go to http://www1.tools.ietf.org/html/rfc4321 etc.,
> 
> The RFC 4321 output is fine.  When I ask for a "generic" draft
> (no number, no extension) like draft-ietf-usefor-useage I get
> draft-ietf-usefor-useage-01 without extension, and the added
> (thanks) text link also goes to draft-ietf-usefor-useage-01.
> 
> If that's working as designed it's okay, but maybe you wanted
> to create a link to draft-ietf-usefor-useage-01.txt (?)

Good point.  Fixed.

> Checking this document:  In its section 1.2 it has a reference
> of the style [USEFOR] at the begin of a paragraph, you missed
> that.  You caught other [USEFOR] etc. within paragraphs.

Ah, yes.  I currently have no way of distinguishing that use of
[USEFOR] from a definition of [USEFOR]; currently it treats this
as a reference definition, rather than a reference.  This is a
flaw.

> At the moment I get an "document contains no data" error for a
> broken link like draft-ietf-usefor-article-format - you created
> the link based on draft-ietf-usefor-article-format-*.txt, this
> is old stuff, you don't have it in your database.
> 
> Checking what's going on, if I try the ID URL for this it's an
> ordinary non-empty 404.  My browser didn't get the fine point of
> the 303 for its GET HTTP/1.0 with the HTML URL.

Oh, I see.  I'll use 302 here.  It's not really correct in this
case, but not harmful.

> Trying RFC 1945:  No 303 in HTTP/1.0.  And also no link to the
> "obsoleted by", pushing 303 issue, now checking "obsoleted by":
> 
> The RFC link (first link before ID and TXT) is rather huge for
> modem users... 2616 was the number I'm looking for, not 2617,
> off by one.
> 
> Looking at 2616, it has "Obsoletes: 2068" without link.  Please
> add links for "Obsoletes: nnnn".  Okay, now I see why 1945 has
> no "obsoleted by", they didn't say "obsoletes 1945" in 2068.

Handling of "Obsoletes: nnnn" added to TODO list.

> Maybe that convention was invented later, or they forgot it.
> Pop 2616 and 303 issue:  "See other" is about POST vs. GET, so
> why do you need this ?
> 
> | Many pre-HTTP/1.1 user agents do not understand the 303
> | status. When interoperability with such clients is a concern,
> | the 302 status code may be used instead, since most user
> | agents react to a 302 response as described here for 303.
> 
> It's not consistent, I get the 303 only for some forms of an
> URL,  And only until you have it in your cache apparently (?)

The difference is that in this case, I don't have a matching
document at all; I provide a page which lists a number of
alternatives... (but see above).

> I just tested rfc1036:  two "document contains no data" for the
> 303, third attempt worked, I got 1036.  Maybe just delay your
> answer until you know what it's going to be, 200 or 404.
> 
> If I try html/1032 (not html/rfc1032) it works without the 303.
> Testing several 103x, always the same behaviour, html/103x gets
> it at once, html/rfc103x runs into the 303 trap.  Later both
> work.

If the document hasn't been cached yet, you probably get a 302
in one case, which works, and a 303 in the other case, which
doesn't.  Once the document has been cached, you get it directly.
This should clear up with using 302 instead of 303.

> Oops, the output "updated by" for 1035 is ugly, numerous links
> all in one long line.  With "errata" at the far right end.  The
> "updated by" meta data before the begin of the original text is
> not forced to use monospaced, you can format it in any way you
> like.  Or fold it after 78 columns if you want it to be part of
> the following monospaced text of the original document (?)

I'll look at folding it.

>> if the browser has javascript, I run a snippet which adds the
>> <h?/> tags, on the theory that most browsers with java-script
>> turned on will also have css which preserves the look-and
>> feel of the result
> 
> Makes sense.  If I turn JavaScript 1.1 on with my old browser
> I'm not at all surprised if nothing works, therefore I almost
> never try this.

Ok, and thanks for checking this :-)

Regards,

	Henrik

_______________________________________________
Tools-discuss mailing list
Tools-discuss at ietf.org
https://www1.ietf.org/mailman/listinfo/tools-discuss