Random (sporadic) author names vanishing in (author-date) styles
I've encountered a bug in the last 2-3 weeks that I had never seen before: vanishing author names. Not all, following no discernible pattern. Say I'm citing Smith (2015), Jones (2016), and Davis (2010). If I group them all together in a single author-date, I might end up with (Davis 2010; 2015; Jones 2016). Or if I try to cite Jones, it'll just show up as (2016).
Oddly, it seems to persist even if I re-add the reference, grab it from a different database... and it's idiosyncratic: it'll happen in one document, then be fine in the next (as far as I can tell). The answer to the obvious "are you sure that in the record, Jones is listed as an author?" is YES. In some cases these are references that we've used successfully many times before.
I'm working around it by clicking "suppress author" (in case it hiccups back to normal and I don't catch it) and then manually entering the author as a prefix, but this is worrisome.
Any idea what's going on? I'm using a custom style, but it's so random/sporadic, and the style has been successfully used for so long, I am completely puzzled.
Thanks in advance for helpful tips!
Oddly, it seems to persist even if I re-add the reference, grab it from a different database... and it's idiosyncratic: it'll happen in one document, then be fine in the next (as far as I can tell). The answer to the obvious "are you sure that in the record, Jones is listed as an author?" is YES. In some cases these are references that we've used successfully many times before.
I'm working around it by clicking "suppress author" (in case it hiccups back to normal and I don't catch it) and then manually entering the author as a prefix, but this is worrisome.
Any idea what's going on? I'm using a custom style, but it's so random/sporadic, and the style has been successfully used for so long, I am completely puzzled.
Thanks in advance for helpful tips!
-
adamsmithI'd want to see the custom style as a start. The citation processor behavior does change -- so if you've been relying on either a certain behavior for invalid code or for ambiguous code, this could be the cause.
-
marionsdI just tried switching to Chicago, and the problem disappeared, so you might be right. Should I post the CSL code somewhere? Or the relevant lines?
-
adamsmithwhole code: gist.github.com and post the link here (you don't need a github account for this)
-
marionsdhttps://gist.github.com/anonymous/308f1bcecad01b7c3c5945e4b4ab3613
-
adamsmithI'd bet on et-al-use-first="0" -- you're effectively saying when using et al, uses the first 0 names, so citeproc is (correctly, strictly speaking) not printing any names for et al type references.
-
marionsdFixed, thank you!
-
Rintze@adamsmith, we should probably change the schema to only allow positive integers for all et-al min and use-first attributes (using the same logic we use for "line-spacing"). I'll create a ticket if you agree. There is no reason to ever use 0 as a value, right?
-
marionsdI have no idea why I used it, btw. Or why it worked for so long.
-
adamsmith@Rintze completely agree, yes.