 Thank you very much and thank you for indulging laptop. It worked and then it didn't and then it worked. So I'm just going to very quickly jump into louder. I don't have a very loud voice, I'm sorry. So this may need with some speaking. Hello, hello. Test, test. I'll just carry on and we'll see. Great. So welcome. And I'll just... Yeah, great. So we're going to talk about what Redis is, the impact of the Commons Clause and what are the next steps as a community, what we actually do and things like that. Some very quick stuff about me, like some sort of what would you call it, hat management, I believe, do it. I mean one hat I've worn in my life is this one. Do you may recognise? Yeah. But in terms of like sensible hats, answering that question sensibly, I'm on the border directors of the open source initiative. I've been a Debian developer, official Debian developer 10, 11 years now. I'm a current Debian project leader. I'm also on the Debian... I'm quite a bit of things in Debian, but one thing in particular that's relevant today is that I'm part of the FTP team and this team, it literally manages the files in the Debian archive, making sure they get put on places where the mirrors can slip them in. They remove old, outdated packages or undesired packages. So we can't have this package anymore and let's get rid of it, or ones that are just like crafty and old like GTK1. It's gone. We remove it. And that's a response with this team. It's a approval of new packages, not only to ensure the namespace of the packages is the same, but also from a legal point of view. So every single brand new package that comes in, it gets analysed for is this pretty software and things like that, and we'll talk about that a bit more later. So reject packages that don't meet quality guidelines, so if they do idiotic stuff, then they're just rejected on those kinds of grounds as well. Including legal grounds. So if a package comes in and it can't be distributed because for whatever reason, then it gets rejected at that point by the FTP team. So I've sort of got a bit of a handle in sort of a few, what do you call it, lots of fingers and lots of different pies. But I'm not really wearing any hats. I'm just sort of, you know, that's just not speaking for anyone on behalf of anyone, things like that. And I'm a sort of, I'm a sort of freelancer person, so I don't have an official employee anyway, so I don't have to sort of do the usual spiel around that. So what's the background here? What is redis? Like, what actually is it? It's pronounced redis, by the way, not redis. Took me like five years to realise that. So it's an in-memory data, it's a disk. If you've ever used Memcash, it's like pretty much the same protocol. If you see what I mean, you can sort of interact with it over Telnet. It's like that one at those kind of things. But in contrast to, say, Memcash, it persists to disk. So what's it sort of in-memory database? It's sort of a no-sequel. It came around, we know, when that kind of thing was becoming a bit more cool or people were kind of pushing against relational databases in some ways. It started in 2009. That's when the first commit is it was actually quite a busy project from day one. Very heavily used in web development, again, because of the same way that MongoDB is like, oh, we can just sort of scale horizontally because we've got an in-memory database. I'm not going to get into that. That's a different question. Things like that. And it's released under the BST from, I don't think the license ever changed. Yeah, BST from day one, yes, definitely. What's Redis Search? So Redis Search is a search engine module that sits on top of Redis. At some point in the last two, three years, Redis got a module system. So you could write external modules. So you could link it with something else that would never go in the Redis core code. Sometimes just for, like, why sort of bloat the upstream repository and all these features. You can just have them as modules and they work just as well. Why not? Things like that. It's kind of cool. Released under the ATPL from the beginning by Redis Labs, who are a company, I think, founded in 2010 or thereabouts. The author of aforementioned Redis Search and launched a cloud-based version of Redis AWS in 2013. Things like that. Great. So, I mean, the timeline here is that the... I first uploaded Redis to Debian in 2009. Things great, lots of users, good feedback. It's pretty cool, used a lot. I know this because if I break it, I get a lot of reports. Yeah, so I think it has quite a few users. It's kind of cool. And that went on for many years. Absolutely no problems at all. And then I saw it had a module system and some external modules. So, I thought, why not package the modules? People are asking for them. Cool. So, I uploaded Redis Search to Debian in 2017. So, as that dates, the previous date implies 2009. No, just great. No problem at all. It was sending some contributions to... As part of the packaging of Redis Search, I sent some contributions fixing some things here. Like, oh, it doesn't compile properly here. Or, oh, this bit's a bit... We're using a newer GCC. So, I just sent some things to fix the warnings and things like that back upstream. Nothing, nothing that crazy. No reworks, almost tiny patches. And in May of last year, I got an email saying, thank you for your contributions that you've made. Hope you find the project useful and can use it. OK? Yeah, cool. Next sentence was, due to a licence of the project users, AGPL, we require that you sign a CLA agreement. Please reply with a signed version of your earliest convenience. OK. Can you elaborate why using the AGPL requires a CLA? I didn't get it. I didn't really understand the mail. So, I just sent it back quite quickly. So, the AGPL has no relevance to CLA, while AGPL is a distribution licence. The CLA is only about contributions to the project. The CLA basically grants Redis Labs a licence with special permission to use the contributions as we see fit. As we see fit. With the CLA, you could just potentially make proprietary forks. It seems somewhat odd. It seems still odd if you have a CLA to have an AGPL project. Like I don't quite see the... Yeah. You want to have a diversified group of copyright holders for an AGPL project or GPL projects in general. If this is right in my ball field, I can honestly say I have no intention of doing anything like that, but I have to ask the lawyers. OK, well, OK, cool. And when we published our AGPL Redis Modules, we were less organised and quite new to doing open source. As time passed, following other companies in this space, Mongo, we decided to adapt the CLA practice. OK? OK, well... Yeah. That said, you actually... LAUGHTER Yeah. Well, they're doing some good-ish stuff in that area. No problem at all. I met some of them, kind of cool. And the stuff I'd done was like super trivial. It was like adding... I think one of them was adding... You know how in C, you have fall-through statements on a switch and one of them was unlabeled? It was those and I fixed one of those. So it was just like, is that even copyrightable? Doesn't matter, but yeah. And then this happens. The Redis Labs Modules, all of them were licensed... Realised from AGPL to what's called the Apache license with the Commons Clause in August. So just to clarify, Redis is not the same as Redis Modules. It's that there's a module system that's part of Redis, but a particular module that's a particular Redis module is not the same as Redis and they don't... Yeah, so Redis, as it says here, Redis itself will always remain BSD, but we decided to prevent client providers from creating managed services from certain modules, including Redis Search. So what is the Commons Clause? There's not a license in itself. There's not a Commons Clause license from a technical point of view. So it's sort of an amendment to a license. It applies a narrow, minimal form commercial restriction on top of an existing one. And so you would apply that on top of... Well, it sort of makes sense to apply on top of the Apache license. If you were going to combine it to that sort of makes sense. In practice, the Commons Clause only adds an imitation concerning fair use. Why do licensing approaches share the same core value making the software available for use by everyone? Well, yes. So what's the problem here? Yeah, we can... Let's take a bit of diversion into the Debbie and Free Software Guidelines. That can kind of frame it quite well. So the Free Software Guidelines were written in... Ian, when were they written? 20 years ago, 20 years ago? Yeah, something like that. I should know this, sorry. They basically determine what constitutes Free Software from a Debbie point of view. And this means whether they go into... This is the test whether they go into Debbie or not. So I'm not going to go through them all, but number one is free redistribution. So obviously Free Software, you should better redistribute it. Make sense. You should be able to make derived works. Make sense is kind of obvious. The interesting ones like license must be not specific to Debian. So this is all around the sort of... You could grant... It's like, oh, this is prominent with your software. But Debian, you can use it for this. It's like, no, no, no, it can't be specific just for Debian. It must be like sort of transitive and it should apply to everyone else as well, which is kind of cool. This was the Firefox ice-weasel issue. The most important one today is DFSG 6. No discrimination against fields of endeavour. So often you'll get things like... Oh, users, anything you want, free software, it's all cool, but not if you're in nuclear facilities and you can't use it in nuclear bombs or things like that. Or genetic research. So you're kind of getting a bit more like... I'm not sure anyone is really like pro-setting off nuclear weapons in this room. But people might be against genetic research here. So there's like a... We're starting on this gradient of sort of moral questions, I don't know, some sort of that sort of area. But non-commercial use is really obviously the one that we're all talking about. Things like that. There's some other similar licences in this space. There's one called the... Oh, you probably can't see that because of the colours. It's called the no harm licence. So it's basically BSD 3 cores. But you're not allowed to use it for anything that promotes hate or violence and things like that. There's a whole bunch of other licences that are like... This is public domain, but you can't use it if you're a Nazi. It's like, well, I'm not pro-Nazi, but yeah. Things like that. But the open source initiative have annotated their own guidelines and under their equivalent clause, equivalent definition part of a licence, we want commercial use to join our community, not feel excluded from it. You know, that's pretty good. And there should be no reason why, whilst money and free software are often sort of uneasy bedfellows, there's no reason why they can't rub along really well. You can make money from redistributing free software or being part of development of it, things like that. And I'm pretty strong, feel very strongly. This isn't a question of definition. So there's... When people talk about the problems with, say, the creative commons non-commercial licence, often everyone's first response is, well, you know what, it's very difficult to define what non-commercially is. Maybe you take my software and make a t-shirt out of it and then sell the t-shirt. Is that commercial activity? Now like these are interesting questions to have over a beer later. But whether a licence should restrict you on a non-commercial basis is not just because one can't define where that line is. It's sort of a moral thing about the licence in general and what should be considered free software and things like that. So when people jump straight to the, oh, it's about, oh, you know, you can't define exactly where the line is. I think that kind of slightly misses the point. And so we'll talk about that in a little bit. So what's the rationale of why you have the commons clause and to begin with? If you read their own words, plan providers have repeatedly violated this ethos by taking advantage of successful open source projects. Yeah, they use the monopolistic nature to derive hundreds of millions of dollars in the, they're talking about AWS, I presume, and that behaviour is damaged and people have gone out of business. So that's a rationale on the other side, things like that. But the, sorry, I'm pushing through very quickly because I realised we were a bit delayed by the laptop things. So the other problem here is that it's, the sort of commons clause and things like that are pushing a certain narratives that the positioning and the way that people, new people might come to free software might be sort of quite slightly changed or if they're thinking of trying to seek employment and they like free software, can it just remain a hobby for them? When you have these big companies re-licensing software, it sort of pushes this narrative of perhaps it just doesn't really work in this space and things like that. So I think that's the other sort of subtle problem with these kind of re-licensing things. So it starts to say, yeah, free software must remain a hobby for you, you can't actually make any money, you can't have a company that's sort of ethical in this respect, things like that. And so by using sort of slightly Orwellian-ish terms like that, if you see what I mean, the source is available, it's not great. And it's also pushing basically the narrative that free software can't be financially self-supporting, which is untrue. I'm standing here based on that, so it's just like, yeah. But these narratives are extremely important, so having them splashed across various bits of news media or very prominent pieces of software in a particular space, like the web development space, does become quite problematic. It might be also be accused of being sort of commonswashing. This comes from greenwashing, even though you've heard that, where people, you know, big oil companies who are, let's just say they're polluting, but they, you know, do lots of nice things or governments investing a lot in education, things like that. It's a kind of like, it sort of looks good to the community and things like that. But calling it a commons clause was kind of quite clever and sneaky, because like commons is, you know, it sort of speaks to a certain other project in this area, things like that. So it's sort of using the sort of moral authority of creative commons, and I don't think that everyone in that organisation would agree with everything that the commons clause is after. Let's put it that way. I can't speak for any of those two parties, but that's probably true. To be fair, to be very fair, the website does say it is best not to call commons clause software open source, because they do agree that it's not an open source licence. They do agree that it's not from the Deviant Free Software Guidelines, one of you, that a licence with the commons clause is not a free software licence from the open source initiative one of you. That's completely conceded on the website, except that on the registered website as of two hours ago, it still says it's an open source. There's a pull request open, since it has some angry people, don't pile on, but yeah. No, please don't, don't. So yeah, there's like, optics are really important, narratives are very important, and things like that. And how you start the conversation is very important. So what are the downsides? Distributions can't ship the software, pretty obvious. We removed Redis Search from Deviant, also from Fedora, and speaking to Nathan Scott from Fedora, we started what's called the Good Form Project, where we took the last non-commons clause version, the ADPL version, and just forked it, and things like that. That's at goodformcode.com, things like that. We're basically going to have no community contributions, because everyone's just going to run away. Here's one. Pile on there, if you like. Just going to power through, sorry. So what else? Where from here? So what do we actually need to do about this? What did we do about it? I have quite a harshest critique of the community response to it, because I think we were right, Morrie, but we just kind of pushed this in the wrong way in many angles, that's kind of interesting. What can economically viable free software look like? Spoiler, I don't know the answer, but what can this actually look like, and work really well? If you want to find out more about that, come to part two of this at copyleft.com from Monday. Tickets are still available. Really important to this. So please come to that. So if there's any questions, please throw them at me now, or I'm not sure around the time. Do we have any questions? That catches us up. We'll have time for one question. How shall we do the critique to the community response? Tools, processes, or places? How should we do the critique? In what way? Sorry. You said that that's the next step, right? Yes, we should look at the way we responded to the re-licensing. So like self-inspection? Yeah, I think so, yeah. That's what you meant. In some way our response was a failure in that it wasn't reverted. Ideally, good form code would just go away, it would go back to being AGPL, things like that. We need to work out why our response whilst we made a bunch of noise, we didn't actually achieve anything. Yeah, so yeah. All right, let's give Chris a huge round of applause.