It’s been 5 years. I don’t think they’re going to change the license to allow distributions to distribute MongoDB more easily.

We should actively be against corporate leeching.

In a world without free software, Amazon will build their own proprietary software for servers that is better than everyone else’s, and will be in the same position. At least with Redis, multiple employees of AWS were core maintainers for Redis. It isn’t like Amazon didn’t contribute anything back. Now that it’s non-free, they’ll just fork it. Again.

All this really accomplishes is making licensing a headache for everybody, which is the main reason people and organizations use free software.

I think free software developers should be able to make money from their software, and money from working on their software. I also think everyone else should be able to, too.

To put it another way, open source means surrendering your monopoly over commercial exploitation.

Additionally, Elasticsearch does not belong to Elastic. Redis doesn’t belong to Redis, either.

Not only companies: https://lists.archlinux.org/hyperkitty/list/arch-dev-public@lists.archlinux.org/thread/OY2DLNWTZOBPAHFE5FSV4Q6AIWGZO6KV/?noscript

Fossify Messages has a release on Github too.

PHP 8.0 is no longer supported so I hope they update the “really, really old technology” to at least PHP 8.1 today.

Most likely. This blog was written in February 2022; support for PHP 8.0 was only dropped in November 2023.

I was half-asleep when I wrote this, lol. Bitbucket dropped Mercurial recently, too. Sourcehut is the only other code forge I know of that supports hg which I really love. Kind of sets a high bar for contributions, but not being vendor locked in is a bonus. And I wish they’d more tightly integrate the subdomains…

I thought Github only supported git, too. Did it support Mercury at some point? I assume this is the last of other VCS support in Github.

You’re not going to convince anyone to suffer inconvenience for something that has no tangible benefit in their eyes. The best you can do is give people the option to contact you on Signal and explain (briefly) why you prefer it. After enough experience, you realize there is no argument you can make that will convince people to care about privacy. The people who join you on Signal either already care about privacy (but maybe didn’t realize it) or value your comfort over theirs.

Personally, I would rather send unencrypted SMS instead of using a Meta-owned service. I don’t want to be part of the network effect keeping people on Facebook. Everyone with a SIM card in their phone already has access to SMS, but few use it if they can help it, so I don’t think I’m contributing to a network effect by doing this. The only MMS client I use is Signal, so anyone can contact me over there if they want more functionality. That’s the only tactic I use, and so far, it has been unsuccessful.

It’s worth mentioning that Android Auto doesn’t work on GrapheneOS due to the privileged access it requires, and will not support it unless it is re-architected. Which phones were you thinking of when you said “compatible”?

I don’t try explaining this stuff anymore. I’m extremely bad at it and I know most people will either not care or not care enough. And that’s fine. It’s easier to let people think I’m a luddite. I prefer to wait for them to ask me why I use technology differently than to preach to them, and most of the time, they don’t ask. It’s always easier to let people convince themselves than to try to convince them.

I tell them to install Signal and message me there, and if they don’t want to do that, they can SMS me. Signal is a better idea if they want faster responses as I rarely check my phone but do check Signal Desktop fairly regularly. I have tried to get KDE Connect working so I can respond to SMS more timely but it was annoying and I gave up.

15 years ago, I thought I wanted to make a game. Turns out, I didn’t.

A few years ago, I sought out Linux. Learning to use it has made me so much more confident and excited about technology. I understood so much more. And yet, it feels like I don’t understand nearly enough. So I’m learning programming so I can start looking through codebases for the projects I use, maybe seeing if I can add new features or fix some bugs that are annoying me. I’ve sort of accomplished that goal for one program. There are also some programs that don’t exist, or don’t exist in the way I want them to that I intend on developing.

I’d like to learn reverse engineering too…

Well, I guess I’m a programmer only by technicality. I haven’t done anything serious and I’m certainly not decent at the art. I’m just curious. 🐇

Did your opinion of the teacher change at all after that?

Honestly I laugh at the people whining about password sharing getting shut down. It was never really intended to be used that way

To quote Netflix, “Love is sharing a password”: https://twitter.com/netflix/status/840276073040371712?ref_src=twsrc^tfw

I haven’t been subscribed to a streaming service for about five years. Well, sometimes I pay for a month of Hidive. I usually just buy the show/film on disc and rip it. A decent amount of the shows I want to watch are old and not even on a streaming service.

Ah, okay. It sounds like you have a physical server, too…you would need to upgrade it yourself if you wanted to use AV1, right? Sounds expensive and annoying…

My understanding is the client needs to explicitly support hardware decoding with the relevant APIs, and Jellyfin probably accomplishes it with FFmpeg. There is no way Jellyfin would be implementing a software decoder for HEVC, but they should have no problem implementing hardware decoders for every platform.

iOS doesn’t even have a software decoder for AV1 yet, but the iPhone 15 Pro hardware decoder is a start. Likewise, only expensive Android phones have hardware AV1 decoders right now. More desktop GPUs are implementing AV1 decoders. But this transition looks like it’s going to take another 3 years (?) to hit every market segment (cheapest to most expensive)… sigh. I don’t have an AV1 hardware decoder on any of my devices either. It’s insane how long it takes for new hardware decoders to become mainstream. Many HEVC patents might be close to expiring by then, lol.

This was a use case I was introduced to directly before I discovered Firefox was introducing support for HEVC decoding.

I use HEVC because it has significantly better compression than older codecs, and many modern devices have hardware decoding support for HEVC.

If it weren’t for iOS, VP9 could take its place, or so the Mozilla developers thought. HEVC and newer codecs like VP9, AV1, VVC, EVC, etc. offer better compression but often at the cost of compute. I imagine hardware decoding evens the scales a lot; I haven’t done any benchmarking myself. I don’t know how much impact the complexity of H.265 vs H.264 has on battery life, if any. Of course, hardware encoding on VP9 is not really a thing (AV1 is ahead of it, even), so HEVC has the edge there.

In a few years, AV1 hardware implementations will hopefully be ubiquitous; that would solve the efficiency and software patent problems at the same time. It’ll probably coincide with the last of H.264’s patents expiring. So on the one hand, I can understand why Mozilla is in no rush to support HEVC.

So I imagine you use a Chromium-based browser for Jellyfin?

I’m on GNU/Linux myself, and personally, I don’t use HEVC at all. I don’t even decode video in my browser most of the time. I’m usually using mpv with yt-dlp. Streaming services like YouTube, Facebook and Netflix don’t use HEVC to my knowledge (being AOM members and all), but I don’t use services that require me to enable DRM in my browser. I don’t know of a service that requires HEVC decoding support.

It’s possible Mozilla will support HEVC decoding on other operating systems in the future. Windows is just the easiest one to start with. It’s worth noting that Chrome’s HEVC hardware decoding support does not support Widevine, the DRM Netflix and other streaming services use. So you won’t see them adopting HEVC in browsers, at least.

The fact that this bug for macOS is a part of the hevc meta-bug indicates that Mozilla also wants to support HEVC decoding on macOS: https://bugzilla.mozilla.org/show_bug.cgi?id=1839107

Chrome supports HEVC decoding on GNU/Linux, so I don’t see any reason why Firefox wouldn’t too, eventually.

In about 5 years, the last patents for the baseline H.264 (AVC) video codec will expire. This means Firefox will finally be able to support decoding for H.264 (the codec most commonly used with the MP4 container) without Cisco needing to cover the licensing fees for them. DaVinci Resolve will also be able to support decoding/encoding on GNU/Linux. Basically, anyone will be able to implement a H.264 decoder/encoder without needing to pay royalties, which means free software programs like Firefox which don’t charge their users will be able to implement it. The codec will no longer be patent-encumbered. See Wikipedia’s debate on whether to support H.264 in 2014 for lots of opinions on this: https://commons.wikimedia.org/wiki/Commons:Requests_for_comment/MP4_Video

However, H.265 (also called HEVC) is the “next-generation” video codec after H.264. The patent pool situation is so confusing and expensive compared to H.264 that Mozilla, Google, Apple, Microsoft, Netflix, and dozens of other companies using video compression technology formed the Alliance for Open Media to develop a royalty-free codec they could use instead of HEVC. As a result, HEVC has seen very little adoption, particularly on the web. Most companies continue to use H.264. Windows 10 asks you to pay $0.99 for HEVC decoding support in the default video player. In fact, Google even announced they were dropping support for H.264 in 2011 because of their “focus on open web principles”: https://blog.chromium.org/2011/01/html-video-codec-support-in-chrome.html

They never went through with it, of course.

AV1 is finally at a point where there are production-ready encoders and decoders. The iPhone 15 Pro is shipping with AV1 hardware decoding support, so even Apple is on board (though only for the most expensive phone so far). So, we’ll see hardware decoding support in more phones in a few years, and hopefully AV1 will see a lot of adoption.

Adding support for HEVC in browsers feels like a step back because it legitimizes the codec and works against AV1 adoption. Many of these companies have resisted implementing HEVC support (aside from Apple) because it positions AV1 as the only realistic option to shift to from H.264. With HEVC in the picture now, it might be a realistic option in the future. Well, there’s VP9 too, but there are some patent disputes over that codec that might make companies sheepish about adopting it. Not that AV1 is free from those disputes either, but they are far more confident about it and Unified Patents is doing great work invalidating Sisvel’s AV1 patent claims (and even some HEVC patents here and there).

My opinion is that the ship has already sailed. Chrome caving in and adding HEVC support last year, albeit only with hardware decoders, was all it took with their 90%+ market share. Firefox’s 3% market share isn’t going to make a dent, and not doing it risks keeping users who need HEVC support on Chrome. I also don’t think hardware decoding support is a big deal right now due to all the older devices without HEVC hardware decoders, but it opens the door to ask for more support down the line. I very much doubt Google is going to do that, though.

Windows users have been asking for HEVC support for years: https://bugzilla.mozilla.org/show_bug.cgi?id=1332136

7 years ago, this was the answer:

Mozilla currently has no plans to support H.265. Our focus will be on AV1.

The reason we won’t support H265 has nothing to do with the difficulty in finding a decoder, or that a decoder source code is released under GPL. Those are trivial matters.

We will not support h265 video while its patent encumbered.

BTW, even today vp9 provides better results than H265.

The conversation changed to, “Firefox could at least do hardware decode support without worrying about patents, right?”

My guess is they’re doing this because Chrome added HEVC hardware decoding support last year.

It’s a good movie, but I find myself completely confused by the driving part at the end. The people who were in the van manage to catch up to him almost immediately, suggesting that the other guy could have just run across the distance to get to the van in the first place, saving everyone all the drama and hilarity.

I thought Wargames was a pretty good one too (mostly). It’s by the same writers.

The NET is…fun…

That’s a nice piece of hardware you’ve got there.

Ah, I can see how that might be useful. I first learned about the command pallette when I needed to instruct everyday users on taking a full-page screenshot on Chrome…it’s far more complicated than Firefox’s method of Right-Click > Take Screenshot. Just another odd thing, lol. Interestingly, Firefox is considering implementing this feature: https://github.com/firefox-devtools/ux/issues/101

I don’t think it would make much sense in my workflow right now but I can see how it would benefit others. Quickly turning on accessibility constraints I’m sure would be very useful. One thing Firefox’s dev tools is desperately missing is search. I get along fine without it, but it would be nice to have.