Doublehanded J/99, prepping for the RORC Baltic Sea Race. I'm building an external analysis layer around Expedition and have run into a workflow question I suspect others share.

What I'm doing. Two cases where I need routing results as data outside Expedition:

• Pre-race: a historical ensemble - the same course routed through X years of ERA5 reanalysis, Y starts per year. That's a lot of routings, each exported by hand. The point is climatological statistics per leg (duration spread, actual sailed TWA/TWS bands, whether a leg has a recurring route choice or just year-to-year noise).
• During/after the race: one routing per weather model, refreshed each cycle, compared against each other. Divergence between models is my uncertainty estimate, and I want it computed and logged automatically rather than eyeballed. Also super useful in post race analysis.

Question 1 — is there an access path I've missed? Today I use File → Export for each routing individually. Is the routing result available anywhere else: a file on disk, an internal database, a COM/API surface, or anything scriptable? I'd happily read it read-only from wherever it lives. I may well have missed an existing option, in which case a pointer is all I need.

Question 2 — how hard would "export all active routings" be? If several routings are loaded, a single action writing them all out (one file each, or one file with a routing-id column) would remove essentially all the friction. Timestamped filenames including the model name would be a bonus - I currently rename by hand, since the export dialog makes descriptive naming tedious at volume.

The broader point, offered as a user observation rather than a complaint: Expedition is built on the assumption that the analysis happens inside Expedition, and it does that job very well. But the surrounding toolchain has changed - several of us now want routings as data for statistics, verification against reanalysis, and model scoring. Making the results easy to get out doesn't compete with Expedition's own analysis; it makes Expedition the trusted engine underneath the rest of the stack.

Happy to test anything experimental, and glad to share what we learn.

Thanks for all the work on the software.

///Peter

Last edited by blur (7/29/2026 2:48 am)