Text and HTML versions of drafts have different boilerplates

[larseggert]

Describe the issue

https://www.ietf.org/archive/id/draft-ietf-oauth-jwk-thumbprint-uri-03.txt (linked from https://datatracker.ietf.org/doc/draft-ietf-oauth-jwk-thumbprint-uri/) refers to the "Simplified BSD License" in its boilerplate.

The HTML version at https://www.ietf.org/archive/id/draft-ietf-oauth-jwk-thumbprint-uri-03.html (linked from the same datatracker page) correctly refers to the "Revised BSD License".

How did this happen?

Code of Conduct

• I agree to follow the IETF's Code of Conduct

[cabo]

Archived-At: https://mailarchive.ietf.org/arch/msg/oauth/UGtcPbhrxntez0CcBiKodiWKX1s

"we submitted both the .xml and .txt files"

The usual outdated xml2rfc at the submitter's end?

The warning against submitting superfluous .txt with .xml needs to be much stronger.

[cabo]

(And the archive link in the message I cited does not work, which is another longstanding annoyance.)

[cabo]

https://mailarchive.ietf.org/arch/msg/oauth/cIeFXccJuk4px3-Iv_wCrQ-s49s/

[larseggert]

The usual outdated xml2rfc at the submitter's end?

I thought the datatracker was re-generating the text version if XML was submitted - @rjsparks?

[rjsparks]

We would have to look, but what I remember is that we don't throw away anything that was submitted - if someone submits both xml and txt, we keep both (think about it from an IPR perspective). I thought we had a path where we would generate txt and warn/error on differences, but maybe that was aspirational and not ever implemented.

[rjsparks]

Until the submission process lets submitters preview the rendered versions, people will insist that we take their text because of potential issues/differences in external reference resolution. It's getting better, but the community trust in the datatracker to not do the right thing with those references has been low.

We really need to solve the "submit a self-contained document" problem.

[cabo]

Until the submission process lets submitters preview the rendered versions,

You say it can’t?
See below.

OK, weirdly, it doesn’t allow previewing the whole plaintext (unless you guess that clicking on HTML, changing .html to .txt, gives you a plaintext view with incorrect text encoding metadata (*)), but in principle the feature is there.

Grüße, Carsten

(*) which you can then fix under View/Text encoding in Safari and probably everywhere else.

[cabo]

(See the broken text encoding at

https://www.ietf.org/staging/draft-bormann-asdf-sdf-compact-03.txt

$ curl -I https://www.ietf.org/staging/draft-bormann-asdf-sdf-compact-03.txt
...
content-type: text/plain
...
$ curl -I https://www.ietf.org/archive/id/draft-bormann-asdf-sdf-compact-02.txt
...
content-type: text/plain; charset=utf-8
...

)

[rjsparks]

I'm going to split this into several issues.

First, idnits needs to stop accepting the older incorrect license name: ietf-tools/idnits#12

The staging area is a mess that needs concentrated attention, but I've sent a ticket to the secretariat to address the webserver configuration issue leading to the encoding troubles.

The "View first two pages" functionality really needs rethought. Keeping the first two pages in the database has sped diagnostics in the past when things broke because the author extraction heuristics went into the weeds, for example, but text isn't the only format anymore, and sometimes the content of those first two pages is not benign. See #4143.

And much of this will change when we take the step in #3577