ini.fatal-error-backtraces

[hakre]

The ini setting so far was only mentioned in the migration guide (new in 8.5).

refs #4886

[hakre]

@DanielEScherzer I saw that you have a long list of changes in PHP 8.5. I only came across it later, so this might be of interest to you. Regarding the change: it concerns the documentation for the core INI setting. I followed the order in the source code since I didn’t have any better guidance. Now with “simpara” instead of “para,” which I had copied earlier. If the INI directive is included in the documentation, that would help me update the reference page for error_get_last().

[DanielEScherzer]

No idea if this should be simpara or para, the entries above and below it use para - @Girgias since I know you have been active with the xml, is there documentation about when to use which?

[hakre]

is there documentation about when to use which?

Yes in the ./docs folder, one of the *.md entries, a section specifically mentions this. additionally there are also discussions in reviews here and there, but its in the docs, I checked that today.

Use <para> sparingly

Use <simpara> in markup (similar to HTML's <p>) in favor of <para>
(similar to HTML's <div>) when there are no block elements (such as
<example> or <itemizedlist> in the paragraph.

style.md quote @DanielEScherzer, and my own write-up earlier, that was confirmed by the ./docs which could also be dated, so this is actual practice:

Rationale:

DocBook has <para>, which includes block and inline
elements, and also <simpara> (includes inline, and
excludes block elements) for paragraphs. 1, 2

Understanding is that <para> should be <simpara> for
all new additions in the php manual where applicable.

Cf. #5429, Tim Düsterhus shared their understanding
during review and Ilija Tovilo confirmed. 3

[DanielEScherzer]

Thanks - I'll merge this in a few days if there are no objections (please ping me if I forget)

[hakre]

Thank you for your feedback and happy holidays.

[TimWolla]

Suggested change

	<simpara>
	The default value is <literal>"1"</literal>.
	</simpara>
	</listitem>
	</listitem>

It seems this error_reporting is the only one that mentions it in the text, because the handling of its default value is somewhat special. The others just use the table above.

[TimWolla]

I would commit the change myself, but it seems you didn't enable the “allow maintainers to push into the branch” option.

[hakre]

Thank you for your feedback, I'm not sure I understand. How I'm receiving it, the default value should not be documented under the entry of the directive because it is not special enough (albeit being default), and despite documenting the value range?

That would appear lazy to me, and in my view collides with what I'd consider a readers expectation under the pretext that PHP directives have default values, to find it documented at the most detailed place in the reference, without the sportive activity jumping across a page to consult a table, then through rows and cells.

And isn't it most likely that the rest of the documentation is also on a best-effort basis and therefore the table is most complete, furthermore with default values, but the detailed entries are of varying degrees and that is most likely the reason, and not a seemingly special case of the one directive you mentioned?

Or is it that it would make the other entries look bad, if seemingly two entries stand out, and therefore you want to prevent that?

Please clarify and also provide the best reference to the in-tree docs you know about, so that I can orient myself on that, too. @TimWolla