Hacker News Reader

Comment by Someone1234 | Hacker News Reader

Worth noting that most of the data MFP uses to power the barcode scanner was community supplied. In fact the majority of their food database was community supplied.

MFP used to be a freemium model, wherein actual value-add was the pay stuff, and the core functionality/community provided stuff was the gateway towards that.

Unfortunately, the app got too successful, and was sold to a venture capital firm: Francisco Partners for $350m. Now they're going to do what VCs do, and try to liquidate out a profit by any means necessary, even if it ultimately ruins the app on the way out.

Open source code is a successful idea; what hasn't really taken off for various reasons is the concept of open source or community owned databases. The closest thing I know of is OpenStreetMaps. A food/nutrition database would be a fantastic open dataset that would largely stand the test of time (while specific products come and go, base facts about specific foods rarely do).

View on HN

Actually there's an open database for food: https://world.openfoodfacts.org/

> In fact the majority of their food database was community supplied.

And it shows. Values for tons of entries seem to be random guesses, with a bias towards lowball, guilt-assuaging numbers. I use MFP regularly, but I need to pad whatever it says by 20-30% unless I'm eating factory-packaged food or measured homemade food.

> would largely stand the test of time (while specific products come and go, base facts about specific foods rarely do)

Really? Not only do prepared products change their formulations plenty often (as you note), but do you really think the heirloom broccolis of 1950 have the same nutritional profile of the industrial-agriculture broccoli of 2020?

Whether you support industrial agriculture or fear it, it directly drives change to the structural and chemical properties of the things it grows. Macronutrients and especially micronutrients are going to be effected by that.

Public nutrient data seems like a really valuable social good, but community-sourced data for this kind of thing is just going to create a noisy, sloppy data set (as it has on MFP).

Under Armor bought it for half a billion dollars from its original creators in 2015. They weren't on target to make back the investment and this sale to VC is an attempt to extract whatever remaining value they can out of it.

What might an open source db look like? It should be publicly readable and forkable at any time, but write access still would be restricted to maintainers. Honestly being able to store your data in git (if it's simple enough) would be a godsend, but I don't see any actual concrete alternatives. Hosting a DB isn't viable.

Another commenter mentioned Wikipedia which is a great model. Importantly, they have both i) automated data-dumps which can be downloaded at any time. ii) an open source platform, which is kind of outside of the scope of the point of just open-source data.

One could imagine a "WikiCalories" site which has all the info that MyFitnessPal did.

> what hasn't really taken off for various reasons is the concept of open source or community owned databases

WikiPedia, and the many wikis around the Internet come to mind.

Francisco Partners is a Private Equity (PE) firm, not a VC. Big difference.

That being said, when it feels like a company is getting greedy— it’s best have your pocketbook do the talking/walking.

Reply on news.ycombinator.com