"Your Data" is Broader Than You Think

NB: This is an old article (2020). Everything may have changed since publication.


On the weekend, publisher Pragmatic Programmers migrated to a new system, which is noticeably faster than the previous one. That's good. But the new version lacks the wish list.

Now, I don't know if it's an artifact of migration and wish list is to be reinstated, or if it was a deliberate decision to drop the feature that probably isn't used by the majority of buyers. But it made me aware, that my "data" is way broader then I thought before.

I've blogged about Indieweb movement at length the last year (in Russian), but even then I mostly thought about my data as data that I consciously create: photos, essays, lame jokes, et cetera. Turns out, my wish list was also useful to me, and I miss it. The same is true, say, about my YouTube watch history and Watch Later list, I regularly refer to it to find some weird video I watched a few days ago.

I don't think that any decision in the chain of events that led to me missing my wish list was malicious, but such is the nature of complex systems, especially web services, that they produce unintended outcomes. That's okay, losing wish list is not a big deal.

This incident made me even more aware that the only data I'm guaranteed to be able to access is the data hosted under my control, either on my own disks, or on the disks of my hosting provider.

What other data I'm not thinking of? That's hard to tell, because this data is produced reactively, as a side effect of using web services normally. Message archives in proprietary services (Telegram, FB, VK, and others), upvoted links to research later on Lobsters and others websites, the set of subscriptions.

I also store my "books to buy" lists at Amazon.com as Wish Lists, which could also disappear at any moment, and in the Cart, which may get emptied. These lists act as my own bibliography of things I'm interested to learn more about, so they do have value on their own.

I'm planning to migrate these lists to my web server as a simple HTML file. HTML files do not require maintenance and also have zero marginal costs.

As to PragProg wish list, I guess I'd have to buy every book they have, 'cause every book published by them that I've read was great.

Published at by mt