Report an issue with an entry

Every entry in CAAIL carries a permanent id, and every card links here with its own id attached. Naming that id is what turns “one of the scaffolding papers looked off” into a correction someone can act on.

Every route below works as it is. If your report is about one entry, though, it is much quicker to act on when it names that entry: entry cards on the Papers, Software, Databases and Datasets pages carry their own Report an issue link, which brings the id here with it.

Two things have no such link, and need none. Rows in a dataset page's Complete data inventory table are listed rather than carded, so quote the accession and say which page you saw it on. Reports about a whole subject theme, or about the library as a whole, are not about a single entry at all.

How to report it

Pick whichever route you already have an account for. They all reach the same person. The first two carry the entry id for you; in Slack you paste it in yourself.

• Open a correction on GitHub Needs a GitHub account. The form is prefilled with the entry id, and the report, the discussion and the eventual fix all end up in one public place with a permanent link.
• Send an email No account needed. Goes to benjamin.bromberg@tufts.edu, with the entry id already in the subject line.
• Say so in Slack No GitHub account needed. Paste the entry id into your message so a curator knows which entry you mean.

What is useful to say

Three things: what the entry says now, what it should say instead, and the source for that. A curator checks the correction against its source before editing the entry, so naming the source is what lets that happen without a search first. A section of the paper, the dataset's own documentation, or the repository's licence file will do.

Certainty is not a requirement. Every report is checked against the source either way, and one that does not hold up changes nothing.

What happens next

A curator checks the report against the source and, where it holds up, corrects the entry through the library's authoring backend. We do not promise a turnaround time: the review path runs through one person, and a schedule we cannot keep is worse than none.

CAAIL's curation is openly in progress. It indexes a curated subset rather than a census and has not measured its own recall, so a reader who knows a subfield well is often the first to notice a specific entry is wrong. That is the signal this page exists to catch.