Lovefield as an enabling technology for "data dignity"

[john dimm]

I made a tool that slices and dices data using SQL queries, and I wanted to make it available for people to use on their own data.  The obvious setup would be to have them upload a CSV which I would then import into a database running on my server.  

But I’m one guy in a garage, so why would anyone trust me to safeguard their sensitive data?  

Lovefield was the answer.  Nothing gets uploaded, just imported into local storage.  So there’s no need for user accounts.  The import into Lovefield is fast, and query speed is excellent.

In a normal server-based solution, all user data gets loaded into a big database running on a powerful machine.  The power is needed to extract each user's data from the global store and deliver it back to them.   It’s so much more efficient to store the data locally, where the database has only the individual user’s data.  

It also means that your data stays on your computer.  It could be a key enabling technology for the "data dignity" project.  Aspects of the user’s data could be useful to aggregate between users (categorizations of financial transactions for example), and it would make sense to pay users to opt in.

With low power comes little responsibility.

The app:

http://www.johndimm.com/breakdown/

The Google Chrome Extension fo use with Mint and Personal Capital

https://chrome.google.com/webstore/detail/breakdown-for-mint-and-pe/npkcbabhgmplkmaaljpcbombbgkggilm

The code:

https://github.com/johndimm/breakdown/blob/master/breakdown/javascript/lovefield.js

 

[Demetrios Papadopoulos]

Being able to easily manipulate data locally, without having to involve a server (whether it is for processing or for storing), is definitely one of the motivations behind Lovefield, and it's very rewarding to see that you and others find it useful for building various Web-based tools.

Thank you for trying Lovefield, and for your feedback.

Happy hacking!

On Wednesday, October 9, 2019 at 2:54:21 PM UTC-7, john dimm wrote:

I made a tool that slices and dices data using SQL queries, and I wanted to make it available for people to use on their own data.  The obvious setup would be to have them upload a CSV which I would then import into a database running on my server.  

But I’m one guy in a garage, so why would anyone trust me to safeguard their sensitive data?  

Lovefield was the answer.  Nothing gets tuploaded, just imported into local storage.  So there’s no need for user accounts.  The import into Lovefield is fast, and query speed is excellent.

[john dimm]

Hey, thanks to the Lovefield team and congrats on a fine product!  

Lovefield means data residing on your computer can be analyzed and presented by a website, and the website owner cannot modify or even see your data.  In a way, it harkens back to the pre-internet days of shrink-wrapped software, when all data was local.  In another way, it doesn’t, because there’s no need to install anything.

“Data Dignity” is all about owning your data, and requiring website owners to pay you for the data you are willing to share.  The strongest way of keeping control of your data is to keep it on your device, not on a server in the cloud.

There are lots of practical issues with that approach, but Lovefield could be the first step. 

On Wednesday, October 9, 2019 at 2:54:21 PM UTC-7, john dimm wrote: