Created attachment 275170 [details] [review]
doc: various improvements

- GSubprocessLauncher exists since 2.40, not 2.36
- more logical order for g_markup functions
- fix short description of GMarkup
- GMarkupParser: specify that some parameters are NULL-terminated.
- g_string_new (NULL); is possible.
- g_string_free (NULL, ...); is not possible.
- other trivial fixes.

Comment on attachment 275170 [details] [review]
doc: various improvements

>  * GMarkupParser:
>  * @start_element: Callback to invoke when the opening tag of an element
>- *     is seen.
>+ *     is seen. @attribute_names and @attribute_values are %NULL-terminated.

I think gtk-doc will complain about that, because "attribute_names" and "attribute_values" aren't elements of GMarkupParser. I think you want `attribute_names` instead. (Possibly, "The callback's `attribute_names` and `attribute_values` parameters are %NULL-terminated.")

>+ * @init: the initial text to copy into the string, or %NULL

@init: (allow-none): the initial text to copy into the string, or %NULL to start with an empty string.

>  * g_string_free:
>- * @string: a #GString
>+ * @string: a non-%NULL #GString

That's implied. Functions have to explicitly document that they allow NULL, not that they *don't* allow NULL.

Thanks for the review, I'll attach an updated patch.

(In reply to comment #2)
> I think gtk-doc will complain about that, because "attribute_names" and
> "attribute_values" aren't elements of GMarkupParser.

The GTK-Doc manual says:
"Use @param to refer to parameters. Also use this when referring to parameters of other functions, related to the one being described."

> That's implied. Functions have to explicitly document that they allow NULL, not
> that they *don't* allow NULL.

For the "free" functions in GLib it's difficult to remember which functions accept NULL (like g_free()) and which don't. Maybe a better solution is to accept NULL in all "free" functions, so it is uniform.

Created attachment 275759 [details] [review]
doc: various improvements

- GSubprocessLauncher exists since 2.40, not 2.36
- more logical order for g_markup functions
- fix short description of GMarkup
- GMarkupParser: specify that some parameters are NULL-terminated.
- g_string_new (NULL); is possible.
- other trivial fixes.

Comment on attachment 275759 [details] [review]
doc: various improvements

I've pushed the commit, it fixes the things pointed above.