• 0 Posts
  • 17 Comments
Joined 2 years ago
cake
Cake day: June 12th, 2023

help-circle

  • I also want to make it clear though, that though I say a lot of nice things in that comment about beeper, they have definitely made some choices I don’t love. Their iMessage bridge situation is just not one of those choices. Moving the iMessage bridge to be selfhosted made it annoying enough to do that it’s no longer an easy consumer friendly way to do iMessage on android, which made apple not persue it as intensely, and effectively ended the cat and mouse game they were in. My iMessage bridge has worked flawlessly for many months, no interrupts.

    They have made a number of decisions which I think are shitty though. Most importantly, their new clients are completely closed source, and can’t be used with any matrix server other than their own. This sucks, because their new clients are easily the best matrix clients I have ever used. Beeper android is just leagues ahead of any other matrix client, especially for it’s heavily bridge reliant usecase. Beeper desktop and iOS also look so so much better than anything else that is available, though I use them less. Beeper iOS has a minimum iOS version of 17 ( Yikes! ) and I wont update my iPhone from 16.2 because I don’t want to lose my jailbreak. on desktop I use nheko because it’s super light on resources, which the old beeper client wasn’t, but the new client kinda is. I’m not a huge fan of all my messages living on beeper servers, and being subject to beeper outages, or potential evil beeper company restructuring ( matt mullenwig is unpredictable!! ). If I could use their superior client with my own homeserver I would be much happier with them

    It seems like beeper deviates from the matrix spec on some things, and nheko has lots of trouble loading images with my beeper account.

    I also wish beeper open sourced their server, because having a nice, easy way to do selfhosted matrix + bridges would be so great. An easy docker container + webui for setup where you choose bridges and give credentials would just be so convenient. I don’t think that’s what beeper has internally right now, but they could build it trivially, and that kind of setup is really how it seems like this should work.

    Currently, on android, if I want a decent experience I’m pretty much stuck with their homeserver and their bridges and their client. All the other matrix clients I’ve seen don’t have the options required to do low priority organization, or mixed rooms and DMs in one feed, or a variety of other things that for normal, non bridge centric matrix, don’t matter nearly as much. You can also only compose message to new people to be sent through bridges on official beeper clients, all of which are becoming closed source, so that sucks.

    I imagine on desktop I could use my own homeserver w/ their bridges and use nheko as the client and have a pretty good time, so that’s good, but I definitely dislike the loss of focus on open source I’m seeing from beeper. It worries me greatly, and it feels a lot like they are going to announce something I’ll hate and can do nothing about. Sofar they haven’t begun building in anti-features, but man, they have absolutely no profit plan sofar that I have seen and it looks to me like them growing one will likely be messy. I would really like for selfhosted fully foss servers / clients to get better and become viable options for the “universal messenger” usecase.


  • Beeper can still do IMessage, just selfhosted, which I think was the right route. Beeper has also contributed hugely to matrix’s bridges ecosystem, and all the bridges can be selfhosted, so there was never any need to give them your credentials, or even use their servers for your matrix host. I like what beeper is doing a lot. They also claim E2EE, and they use matrix so that’s entirely possible ( likely even ), but it’s also entirely possible that they can snoop on your messages as their version of the matrix server is closed source. They are also working on on-device bridges and have already released one for signal so you never have to send your credentials to any beeper servers. Beeper solves a ton of legitimate problems for me with stuff like client compatibility across platforms ( I can use any terminal to send iMessage now. That’s wonderful. ).

    Eric’s projects consistently strike me as cool. It doesn’t seem like he really wants to do them himself though, it seems more like he just wants them to exist, and the only way that happens is if he does it. I assume that with the beeper situation he sold it because he assumed matt mullenwig would keep it going and the problem of unified messaging is solved. Then matt went a little crazy and that was less cool. I don’t know if there were signs of him being like this before the wp engine stuff, I only started paying attention to him after beeper sold to automattic.

    There are a lot of problems in my life that I want solutions to that I think could be sold as products, but I don’t particularly want to run a company. I don’t think products will ever be developed which solve some of my problems. I could totally see myself doing what Eric has done here.

    It seems like Eric bringing back pebble is just him wanting a good hackable watch. I don’t see why it needs to be any deeper than that. He doesn’t seem profit motivated ( otherwise why would you wade so deep into open source ).



  • Mobile data you pay a service provider for and link all of your information to ( address, name, etc ), and can be used by one company to track your location at any time with very high accuracy as long as you are near 3 cell towers. Public wifi gets no information about you other than your MAC address and that you’re currently within it’s range. There is no central body that can track all your movements. You could, theoretically, buy prepaid data plans to minimize the info they know about you, but then you have to buy a new one each month, and there’s STILL one company tracking all your movements each month, though they don’t really know who YOU are. They could still do traffic analysis to figure that out.

    It’s not that it’s less secure, it’s that it’s worse for privacy.

    Also, messaging over SMS / MMS is awful for security, which I lump in with the rest of this conversation. https://youtu.be/wVyu7NB7W6Y


  • To be clear, I have a phone number, but I do not WANT to have one. Most aspects of my life I have removed my phone number from. There are still a few services ( like signal! ) which requires one, and I cope. Cellular data is also something worth avoiding, from a privacy perspective. It is very possible to live a life where you’re never very far from wifi, especially in a city. I do not currently do this, but would love to one day.





  • Signal requires a phone number on setup.

    Also, matrix has bridges, which alone make it worthwhile for me. They, of course, don’t help privacy, but they are so so nice for convenience.

    Matrix is definitely slow though, and a grand majority of the clients are heavy terrible buggy electron apps. There are a few good ones ( nheko and the new beeper clients ), but even they have some rough edges.

    I still use matrix all the time and love it.

    If max privacy was the goal I think simplex looks wonderful. No required info for sign up, no way for them to possibly collect any metadata ( because there are no identifiers sent over internet for anyone at all ), E2EE, and decentralized.






  • I use obsidian but I wish there was an open source notes platform that could do what I want:

    1. Excalidraw support ( or similar ) with PDF import and annotation support ( this is achieved by a plugin on obsidian )
    2. Vim mode
    3. Markdown for everything

    I have tried so many notetaking tools and the closest I ever got was using xournalpp for PDF annotation and drawing, then writing plain markdown in helix / neovim, with a live markdown rendering pane on the side. Was just too clunky though.