Skip to main content

Haskell Book Index Mistake

I have started to migrate the Haskell Books data for use with FeedPipe. FeedPipe is designed for a very specific purpose, while the website software is much more general. There are a number of changes to make, but both use Markdown and YAML to specify content and metadata, so migration is not difficult. I completed the initial migration but plan on reviewing all of the content from a FeedPipe perspective.

During the migration, I discovered a mistake in the index. The year of publication of Category Theory For Programmers was correct on the book page but incorrect in the index. This discrepancy is possible because the data used to build the index is separate from the book page data. The website software cannot (yet!) render a page using the metadata of other pages. With FeedPipe, the context for an index page includes the metadata of the items being indexed. This makes content easier to maintain, reducing the likelihood of such mistakes. Note that this particularly mistake is impossible to make with FeedPipe because the year of publication is only specified in a single location.

Author

Travis Cardwell

Published

Tags