Plaintextify the HTML introduces spurious asterisk bullets for some lists

[jgscudder]

Describe the issue

I noticed that the "Plaintextify the HTML" rendering for at least one list in draft-ietf-idr-long-lived-gr-04 (Section 3.1) inserts spurious asterisk bullets. "HTMLize the plaintext" is fine, so is txt of course, and so is native HTML. PDF is broken in the same way as the Plaintextify rendering, I assume the one is used to produce the other.

(04 is where I noticed it, but when I look back at 01 which is the first version for which there is a native HTML rendering, unsurprisingly the artifact is there too.)

Code of Conduct

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

[larseggert]

CC @martinthomson

[martinthomson]

For reference: https://datatracker.ietf.org/doc/html/draft-ietf-idr-long-lived-gr-04#section-3.1-4

Already fixed upstream at https://github.com/martinthomson/rfc-txt-html/blob/2b0a8173ac8d262e72aa242db00dc211d8c05547/txt.css#L219-L221 (other related rules change as well).

[martinthomson]

BTW, that section of that draft has a very strange layout. I don't know whether this is down to the draft itself or what xml2rfc does with it, but it looks like the line spacing is wide at the start of the section, but the last point "Long-lived Stale Time" has no spacing at all. (Given the contortions here, I'm pretty chuffed that the styling--other than the bug mentioned--is perfectly consistent with the text output.)

[jgscudder]

On Mar 12, 2023, at 9:15 PM, Martin Thomson ***@***.***> wrote: BTW, that section of that draft has a very strange layout. I don't know whether this is down to the draft itself or what xml2rfc does with it, but it looks like the line spacing is wide at the start of the section, but the last point "Long-lived Stale Time" has no spacing at all. (Given the contortions here, I'm pretty chuffed that the styling--other than the bug mentioned--is perfectly consistent with the text output.)

Huh, so it does, at least in the txt rendering. The HTML rendering looks as intended, but the other renderings lack vertical spacing to set off “Long-lived Stale Time” as you point out, and I agree that’s mighty peculiar. I’m puzzled that even though the HTML looks good, “plaintextify the HTML” exhibits the same pathology as the txt, which I presume was not used as input. The txt rendering appears to have been busted since version 01 which must have been the first one I submitted in XML instead of txt form. I know I wasn’t hand-repairing the txt, I’d remember that, but something about the renderer must have changed and I didn’t notice because I was reviewing the HTML output, not the txt output. Admittedly my XML is pretty tortured to produce even the desired output. The reason we used this hard-to-generate format is lost in the mists of time, but considering the draft is in the AD’s queue I’m not inclined to change it now. I noodled around with the XML a little to see if I could fix the vspace problem, and failed — I guess it will fall to the RFCEd to undo my sins. :-( Anyway, thanks for pointing it out. —John

[martinthomson]

@jgscudder you might be able to use a definition list to alleviate some of the problems, though I suspect that the figure will continue to be a nuisance. Using vspace might use make things worse. Of course, RFCEd experts will likely be able to help.