This was primarily a preparatory release for an upcoming rewrite of the import engine. The goal was to get the system running more efficiently, more accurately, and to prepare it for the upcoming modifications, which are extensive.
Job Import Accuracy Improvements
Jobs are now processed in bulk asynchronously instead of in serial as they arrive. This backend processing change improves the accuracy of the import process and reduces errors that result in dropped jobs. A backend switch was also made in what data the import process uses to index jobs. This has decreased the load on the import server.
Separate values for Page Title and Page Heading
Members may now choose to have a different value for the site title (which appears in the browser tab) and the site heading (which appears at the top of the page). Previously, there was only Site Name, which was used for both. If they wish, members may still use the single site name property for both title and header.
Markup was added to jobs to bring them into compliance with Schema.org’s job, location, and organization (in our case, companies) properties. Schema.org’s syntax is endorsed by the US Department of Labor. Previously, this data was only on job detail pages.
Performance and Logging Efficiency Improvements
Database indexes were optimized and redundant issue logging packages were removed.
New production servers
To improve performance, we deployed onto new production servers using an improved deployment process. The old servers have been retired.
Major Bug Fixes
Job Listing Title
A previous modification to add an “All Jobs” link to the site broke the dynamic nature of jobs listing titles. That has been fixed, and job listing titles now reflect the search parameters of the listing.