Last month, when I'd been messing with my Goodreads (GR) export file, I realized something I should have noticed ages ago: the export does not contain "date started". That will not do.

On GR, it works like this: if you add a book to any of your 'shelves', i.e. you mark it as "to-read", "currently-reading", or "read". If you do all of those, GR marks the date you first added it to any shelf (called "date added"), the date when you started reading, and the date when you finished. I made sure to keep track of when I started and finished those books, which means I had over 700 date pairs there.

Then I noticed that in the export, what I thought was the "date started" and "date finished" pair, is actually "date added" and "date finished. That is, GR lets you export the date when you first added that book and finished reading, but not when you started reading. No wonder why: Amazon loves its walled garden and wants to increase switching costs as much as it can. I guess it's my bad for not noticing sooner.

However, since my reading records started way before GR and I intend to keep them even after GR is gone, one thing is clear: I will never again use GR to track my reading habits.

Since I'm not giving up on my data, I decided to get it out by hand if I have to. Fortunately, I've been able to script it; more on that in a different post, perhaps.

(Update: the post is here - Rescuing your data from Goodreads.)

Where next?  Storygraph?

SG is a popular GR alternative. It boasts a great recommendation engine, which I don't care about. It lets you import GR data, and export contains all that it should. It contains plenty of other features, like "buddy reads", which look fun but can't be exported, so I won't be using them. And there is no API.

But the GR CSV is the only kind of import they support; so even if you, like me, do have the 'started reading' dates, there's just no way to get them into SG. Not sure I want to continue using a webapp that can't completely do the most basic task I use it for.

And the team behind SG is very small; so even if I charitably assumed they set out to build a friendly and open system instead of yet another walled garden, it's clear that interop features aren't a priority and with such a small team, they won't be available any time soon.

So, what do I do?

At this point, maybe it would just be easiest for me to stick to my own evidence system, just like I use for my games or physical book purchases. But I like having a book DB where I can simply find a book, look at the cover, easily mark the start and end of reading.

The social aspect also isn't negligible. I like seeing what my meatspace friends read; and apparently, so do they. But maybe I'll still use a system of my own, and expose an RSS feed? Or one of those fancy JSON ones, specifically for book tracking?

Perhaps. We'll see. Definitely won't be returning to Goodreads though.