• 0 Posts
  • 21 Comments
Joined 1 year ago
cake
Cake day: June 5th, 2023

help-circle






  • Re: port-forwarding, I used traefik as a reverse proxy and that worked well (having a single domain cert instead of per service DNS is another layer but it’s just obfuscation), but it’s always a risk. I finally started using Tailscale after hearing about it for years and it is actually very good and deserves the hype. I had meant to setup wireguard myself but this is a lot easier. And if you don’t want to use tailscale server, you can run headscale (on a cheap VPS?) instead.



  • Responding separately to the license bit…

    MIT licensed projects (like the libraries, etc.) you’re using allows it to be packaged with products that are governed by other licenses. MIT is a very permissive license and while I’m not advocating for a more restrictive license, I wanted to point that out.

    The other point that @Perhyte@lemmy.world pointed out is also a bit confusing about the conditional licensing. Can a commercial entity use this software as a MIT licensed software as long as the flag is set properly? If so, it would be helpful to delineate what functionality is restricted. I haven’t seen conditional licensing based on run time settings before so I can’t speak to that but it would concern me to use it in any commercial endeavor even if I agreed to the business license.

    I hope you’re taking these comments in the spirit they’re written, asking for clarification and providing feedback to help and not just a critique aimed at a takedown. Cheers!


  • Thanks for the reply, being able to see the data collected and then click delete is great. Does Bespoke keep all the data on its server and allows the customer to get aggregated results or does the customer get to download the raw data? If it’s the latter, the delete functionality becomes… less functional.

    Unsolicited advice from internet nobody: I think it’s great to allow any kind of transparency in a very opaque industry, thanks for doing that. Since you don’t/can’t control the data usage after collection, I think leaning too hard into transparency and alluding to data sovereignty/privacy (by presenting yourself as an alternative to exploiting user data) might create unrealistic expectations (like it did for me). This is inviting unnecessary critique and distracting from your main message.


  • Hello Afi, it’s great that you’ve decided to open source your software and want to drive transparency. If you’re willing to share, I have a couple questions.

    While it’s great that the data collection can be transparently observed, it’s not just the actual collection that’s problematic. Data ownership, protection, storage, usage, associating with other data sources, creation of personas (accurate or misleading) is concerning.

    • Do you have plans to address those in a transparent manner as well?
    • How about user choice in opting-in to data collection?
    • Or an observable deletion of the data (once collected) upon user request? In all the places the data is shared to?

    Lastly, and unrelated, what made you choose an MIT license instead of say AGPL that would better protect your product/company while still being open source? I like the MIT license a lot, just curious.

    Wish you the best in your open source journey!


  • Not what you’re asking for and others have provided a lot of options… but if you still want to support authors and get DRM-free audio books, check out https://libro.fm. It works like Audibe credits and easy to pause and resume membership (more cost effective than buying books retail and you can buy extra credits). They also support a local brick and mortar book store of your choice with your purchase.