Jason (00:02.734)
Hello there, this is Juan Carlos and we are listening to Linux for Everyone in Caracas, Venezuela. Welcome home.

Jason (00:36.248)
Hello, my friends, and welcome back to Linux for Everyone. I am your completely out of hiatus host, Jason Evangelo, and this, as you may remember, is a show about the exciting world of desktop Linux and open source software and the community creating it. I have missed you. You know, the last episode of Linux for Everyone, the last proper episode was recorded November 2021.

And when I stopped to think about how much the world has changed since then, mean, the steam deck launched. We were still very much in the midst of a global pandemic. The ongoing war between Russia and Ukraine hadn't even started yet. Chat GPT wasn't even a thing. And now it's part of so many conversations across so many industries. And I've changed a lot.

I honestly don't remember if I told you the show was going away or going on hiatus here on the actual show or if I had just announced that on social platforms. What I need you to understand is that I was struggling with burnout. I was burnt out, I wasn't struggling with it. I was burned out. And it had been an ongoing struggle for a while. That's why

I had tried to sort of reinvent the show with additional co-hosts and Shickle and Jerry, I will forever be grateful to your contributions during that time. You did help me keep the show alive, which I very, very much appreciate. But yeah, in a way, my drive and my creativity was also crushing me. I was trying to grow everything too fast. was trying to...

Launch new brands, know, launch tech for everyone and make it Linux and I make FOSS and all these things were happening. So many wheels were spinning, so many plates were up in the air. And at the same time I was, I was trying to make a full time living off of the Linux for everyone brand. I was trying to produce enough videos and enough shows and retain sponsors so that I could keep putting food.

Jason (02:59.456)
on our table and the pressure of that it just got to me and that's why I decided I have to get a quote unquote day job. have to find something that is going to be financially stable and something where that financial stability wasn't dictated by

by a fickle YouTube algorithm or a video popping or a sponsor signing up for a few months. And the happy ending here is that I did find that work and I found it with Mozilla. For those of you who don't know, I have been, since last April or May, since April or May, 2022, I have been the marketing manager for Thunderbird. And I know you've all heard of Thunderbird, the...

open source mail client that's been around since 2003 and in a previous incarnation called Mozilla Messaging. And so I'm also doing a podcast called The Thundercast. I host and produce that alongside my colleagues over there, Ryan Sipes and Alex Castellani. Now let me read you the description of The Thundercast, an inside look at the making of Mozilla Thunderbird and

community-driven conversations with our friends in the open source software space. Doesn't that sound familiar to you? Where have I heard about that basic premise before? so fast forward, well, fast forward, sure. Fast forward to a few weeks ago and I attended my first Linux app summit, LAS in Brno, Czechia. I have to tell you, it felt like a family reunion.

There was such a warmth and such a passion for Linux and open source there. And it's what finally pushed me over the edge to to truly, truly missing doing this show, doing Linux for everyone. And I also got really fired up about a particular topic that I had never thought about before. It was a talk that was given at LAS about software sustainability and the man giving it. Joseph was with KDE, is with KDE.

Jason (05:14.728)
And I'm not gonna say too much about this upfront because he is our amazing special guest for this episode, episode 58. But the bottom line is all of these things coalesced into this burning desire to start up the show again, to power it back on and use the platform to say things that need to be said and share things that need to be shared.

I now believe that in the interest of keeping Linux for Everyone healthy and thriving and alive and continuing, that it's best not to monetize this brand and this show. there won't be any more Patreon. For the time being, there won't be any more merch. And I'm not going to kill myself to stick to a schedule, you know?

When I feel like I have something important to share or or something awesome from the Linux and open source community to share, I will press record. I'll invite people on and will release a new episode and I'm also changing the license type of this show. It's always been under a traditional copyright and it is now Creative Commons and it's specifically a Creative Commons attribution non commercial share alike license.

I'm doing that because I want to invite you, the community who has always supported this show and always believed in my journey and the, you know, the enthusiasm that I have for sharing this stuff with people. yeah, I want to invite you to start participating somehow. Some type of community driven Linux for everyone show.

I don't know what that looks like yet, but I'm very open to ideas, so let's talk about it. You can find me on Mastodon. You can also email me Linux for everyone at PM dot me, but check the show notes and I'll have links to my Mastodon account and that email. Anyway, Linux for everyone is an important platform. And I want it to continue even if I'm not always the one behind the microphone. There is so much.

Jason (07:33.88)
collective knowledge and collective passion out there from everyone who listens to the show and everyone who participates in the community. And I want to get all of that revived and I want to somehow try to harness all of these great things that everyone's working on and have the communities help for doing that. And so that's why I think that changing the license type and basically demonetizing the brand

and inviting communities to participate somehow is the most important thing going forward. So with that necessary monologue out of the way, let me turn the spotlight to my conversation with Joseph Devoguis. And if you're anything like me, this could very well be a shocking, eye-opening call to action. Enjoy the conversation and I will come back on the other side with some closing comments.

I am absolutely thrilled that you have chosen to be part of this. I don't know if I want to call it a rebirth, but we can definitely call it the end of the Linux for everyone hiatus. Go ahead and just introduce yourself real quick and tell everybody what you do.

So my name is Joseph Devoguis. I work at KDE in the KDE Eco Project, which is a project focusing on sustainability in software design. I've been working for KDE for about almost two years now. So I started in July, 2021. In my former life, I was a researcher.

working in experimental and theoretical linguistics. So a very different field. But I've been a long time FOSS enthusiast, coming at it mostly from a sort of political and social angle. People who know me, who aren't software developers think that I'm a super user. People who are software developers know that I'm not. I just like to...

Joseph (09:41.996)
I like to change configuration files and things like this, that's, yeah, but I believe strongly in free and open source software on many levels, but coming at it from the social and political side, think.

The social and political angles of FOSS seem like they can be a little bit divisive and heated. Has that been your experience?

So it hasn't been, but there might be a few reasons for that. One, I have avoided social media or the big social media platforms because of my politics, because of my belief in privacy. I'm not active in those channels, so that's not the environment that I'm talking about these things in. And in Berlin, I think there is a...

a long history of alternative ideals about how society can be organized differently. And there is a strong free and open source software, alternative digital society, community here, communities. So I feel that I've been in friendly waters. I also, maybe another thing is I try when I talk about

the political and social issues related to software. I often try to find ways to relate to the values that people already have, trying to connect what I think is there when it comes to digitization, but many people don't recognize the connections. And by making those connections, usually it makes it less, it's less of an effort to understand where the political and social.

Jason (11:27.278)
It is quite a pivot going from a theoretical and experimental linguist to a very passionate evangelist for software sustainability. I want to, before we go forward, I want to hear about your origin story. That's kind of a tradition here on the show is finding out what makes people tick, like what their first experience was.

that they remember anyway that made them fall in love with Linux and open source.

Yeah, I come from, I was a digital skeptic until maybe 2009. And what I mean by that is I tried to avoid computers and thought of them as something that was taking away experiences in life. I felt this shift to screens and shift away from real life interactions. I was skeptical that those were good things.

very early on and I grew up in that in the era when the World Wide Web was new in the early 90s and I remember going from, you know, connecting to message boards, talking about music, that was what drew me to the internet and suddenly having pictures show up.

It went from a black and white or a green and a green colored screen with just text and sort of links to suddenly having images and then later video and things like this. So I saw that transition. remember what it was like before the internet became integrated into every aspect of life.

Jason (13:17.494)
Yes. Okay, so I was chuckling about that because just a few days ago on the Thundercast, which is the Thunderbird podcast that I host, we were talking about what Ryan called the before times. And he said something that really, really stuck with me. And he said, do you remember when we used to go to the internet? Like after dinner?

You know, and his dad would say, well, where do want to go, son? And he'd say, HotWheels.com. You know, you intentionally went to the internet and then you left the internet. And it was such a different experience than just being surrounded by the internet 24-7.

And having a digital correlate to every aspect of our analog life. So I'm fascinated by those things and they were always a part. actually, so before my studies in linguistics, I started late in my career in linguistics and I started learning programming and that demystified the computer. And suddenly I saw how

how people determine what computers do and how we can influence it and shape it. And that skepticism became a desire for autonomy. And very quickly, I landed in free software. This happened within a year. Suddenly, I was very excited about Linux, started exploring.

various tools and this was mostly done outside of the university context, but I was getting a lot of input through some of the tools we were using for corpus analysis, where you run scripts over large databases of language to get natural input for a study you're doing. Things like this, running statistics, statistical programming, which is a different kind of programming, but it's still a type of programming.

Joseph (15:26.894)
and giving me a sense of, I don't know, self-determination in the digital space that led me further and further away from proprietary software because that did not give me any sense of autonomy. For all of the ways that it could enable certain behaviors, it also felt like it was often accompanied with something that was exploitative or abusive. And free software was a space

where it felt that computing was working for me and for the communities that I was involved in in a different way. And then there is the whole community aspect to it, which is hard to explain to someone who hasn't experienced it, but there are real people developing the software and you can talk to these people and you can influence and shape the development of the software. You can contribute to it yourself.

You can contribute to it in many ways if you're not a programmer. You can help with... And you can contribute images, you know...

Thank you

Jason (16:33.518)
Yeah, back in the first few months that I had started using Linux, and this, I'm not sure if I told you at LAS, but this was only back in 2018. I was only about two or three months into it, and my love for it and free software and the community like exponentially grew because every time I would reach out and...

and have a question about Pop OS or about the development of, you know, the vocal app or something with elementary or, you know, some hard, some piece of Linux hardware, there was always a response from somebody or multiple people immediately. And there was always like this, this, this comfortable dialogue that came with it. And that blew my mind because I was coming from a world where I was covering,

Windows and Nvidia, graphics cards and Asus. If you could get a reply, always came with all of these stipulations and this PR overlords looking over the shoulders. So it was just a breath of fresh air, really. And that feeling of community has always stuck with me.

I mean, it's why the tagline for this show has always been exploring the thrilling world of Linux and open source software and the community creating it. I'm sorry, gang, I haven't had a lot of time to gush and rant and have all these little happy thoughts with you over the last year, so I've got to get some of it out.

Yeah.

Joseph (18:15.406)
Yeah, that's, mean, things changed. Edward Snowden changed a lot. That was a moment of, became, it went from something that I felt personally inspired by where then it became, I feel like I need to share this experience that I had of gaining control over my digital life.

in an environment in which it suddenly felt like we have no control and power dynamics have shifted so far from what I think, remembering back to the 90s and this idea of the internet being a democratizing force. So going from this democratizing force to something that felt like the power structures shifted

dramatically into this very not democratic form, suddenly made all of this something that I wanted to share very intensely with other people. And I got involved in various communities in Berlin, started a group, which I mentioned to you called Slow Tech Berlin. It's inspired by slow food and the idea that, you know, cheap and fast aren't the only two values that we should

evaluate technology by, that there are other values that might mean things take perhaps a little longer, they might go a little more slowly, but they're worth it because of the other values that we put into the technology that we're using.

So what you're saying is thin and light laptops aren't the only thing that matters.

Joseph (19:58.489)
Yeah, sure. Absolutely. Yeah,

You know what I'm saying though, like, people go, okay I need this because it's 2mm thinner than the last model and it has two more cores.

That means giving up all personal privacy and contributing to the tsunami of e-waste. Then no, it's not. That's not worth it, in my opinion. And where that line is is a discussion I think we need to be having. I don't want to say I have the answer to where the line is, but I think we haven't been having that conversation.

I mean, yeah, this is exactly where the meat of this conversation begins, I think, because your talk at LAS was shocking to me. It was shocking. I mean, it was a wake-up call to me, and I don't know how many others in the room felt the same way, but it has consumed my thoughts ever since LAS. You know, I was thinking about the backlash that happened.

surrounding Bitcoin's energy usage, right? And cryptocurrency mining in general, know, everyone was up in arms about the environmental toll and the power consumption that was consumed by just a single Bitcoin transaction or, you know, the fact that the mining operations in this area were the, I don't know, some horrific number of

Jason (21:25.644)
you know, the size of a small country's annual power consumption or something. That was fair. That was perfectly fair. I'm not defending that at all, but what I'm getting at is I've always thought about things like, my, is my PC power efficient? Is my Linux distro power efficient? But I never, until your talk, thought about things like, well, is Audacity power efficient? Is

Is Jitsi power efficient? Is OBS power efficient? And I'm talking about all the three things I'm using at this moment to record this show. Is Thunderbird power efficient? What is the difference if I use Firefox versus Chrome? I never thought what does software have to do with environmental issues, right? And that's exactly where you step in.

And this is, I'm very happy to hear that it started a thought process, which is hopefully going to result in more inspiration for measuring, for getting data. Because those, questions you just asked, those are empirical questions. We can measure how much energy consumption is driven by OBS, is driven by JITSE, is driven by...

other software products. And until we have data, we can only speculate. The project that I was working for, the Blower Angle for FOSS, one of the things I liked in particular about the anchoring to the Blower Angle Eco-certification methodology is it provided a clear methodology. It gave some, you know, a framework to start talking about these things, at least in the desktop software domain.

I'm glad to hear that it left an impression and I'm glad to hear that you are, that what I think are important questions we need to be asking and will lead to, I think, important steps being made to getting some data, to getting some answers so we can start thinking about how we improve things, how we make them less environmentally harmful, how we drive down energy consumption and keep hardware in use and find the right balance between the

Joseph (23:42.892)
all the benefits of digitization and reduce the drawbacks.

Can you run down some of the, I guess the headlines, the highlights that would convince people that this is a discussion that we need to be having?

So one example that seems to really impress people is the way when you scale up energy savings, which in software, when we release a piece of software, it gets installed on millions, if not billions of people's devices. And the example I gave in the talk is one CPU second reduction in a piece of software.

which is the equivalent to about a 10 watt second savings, which is nothing. But once you multiply that out by even a modest number of users, I think the example was 1.5 million in the talk I gave, it can quickly add up in a year to the power savings of a small city like Turin or Lisbon. And that kind of scaling up leaves an impression because it's...

easy to see how once you multiply by the number of users, which is easily in the tens or hundreds of millions for many software products. These make real big

Jason (25:11.086)
So my understanding of a CPU second is Is is essentially it's not it's not a literal second, right? It's just a unit of measurement that That tracks How long it takes a CPU core to do a transaction and a transaction? From from what I understood is something like if you know, imagine you're shopping on amazon.com You add something to your cart. That's a transaction

you add a shipping address, that's a transaction. Is that generally how that works?

I should go check my notes to make sure I'm saying it correctly, but I believe it's for how long this CPU is active for, but I don't... I think it actually literally is a second. But this example is not my own. This comes from Detlef Tomes, I believe is the name of the programmer who gives a course at the Hasso Plattner Institute. Important to understand though is that...

The use of CPU correlates with energy consumption. is, I think, you can look at the energy consumption graphs, for example, for ocular, which have been released as part of the transparency requirements of the Blue Angel label. You can see the CPU and the energy consumption correlate. So when CPU is engaged, when it's being used, you're also consuming energy.

So I know that we have a lot of developers out there listening and I would love to know what is the effort and the time involved in actually taking your software and optimizing it and shaving off those CPU seconds, know, even one. What is shaving off one CPU second from your software's transactional time? Like what does that entail? I would love to know that.

Joseph (27:01.792)
I saw a comment recently on Mastodon, is eye-opening to think about is how fragile software is that a couple of lines of code can be a difference between 5 % CPU usage or 100 % CPU usage. What? Right? And it's just, can easily, you can easily do something in a way that results in, you know, really inefficient code. And software is extremely fragile in this sense. And you see this, know, bugs are introduced and suddenly

things are crashing and you know freezing and things like

I right, I mean, I remember, and this has of course happened multiple times, and I know it's easy to pick on Windows, and I'm sure this has happened with Linux distros, it's just, my field of vision involves seeing this happen to Windows a lot in the news. Imagine a Windows update comes out, it's, know, patch Tuesday or whatever it's called, and there is some bug.

that happens where when you're booting up for the first two or three minutes, your CPU is pegged at 100%, nobody quite knows why. Multiply that by 500 million users. And it's probably way more than that, right? I think it is. It's probably way more than that, right? It's probably like in the billions. That's preposterous. That's bananas to think about when we're talking about how much collective energy

that mistake consumes. And the, I guess the ecological footprint that that could leave.

Joseph (28:34.894)
One area that, because it also overlaps with my interest in privacy and privacy oriented software, is all kinds of data mining and serving ads and analytics and the ways that all of this follow a economic logic, not a technical logic. These are often not...

processes that are necessary for the software use, but for the, you know, the profits of the companies behind them.

Can you talk a little bit about the research that was done regarding advertising on cell phones in Europe?

Yeah, so there was a report. It was for the European Union. It was like a policy report. So, okay, the title is from 2021 and it's called Carbon Footprint of Unwanted Data Use by Smartphones, an Analysis for the EU. And what they looked at was the environmental costs of data mining and advertisements that users can't turn off.

That's why they call it unwanted data use. And they can't opt out of. And they looked at the energy consumption for, I believe the number was 60 % of EU citizens. And that number comes from a poll that found 60 % of smartphone users would turn off ads if they could. And they analyzed how much energy is consumed just from that 60 % of unwanted data use.

Jason (29:50.498)
Right, they can't turn off, okay.

Joseph (30:15.47)
And the numbers are staggering. This would be equal to the carbon footprint of between 370,000 and 950,000 EU citizens in one year.

And that's looking at things we don't think necessarily think about, It's the servers, the cellular bands, and the networks working. like I said, the ripple effect is insane.

the network.

Joseph (30:42.638)
These are the kinds of things that most users aren't going to necessarily see in their own devices' energy consumption. This is outsourced energy consumption. It's going through networks that are being powered outside of your device. I think that's one of the ways that a lot of the environmental costs of digitization get hidden from everyday users. They don't see it.

Yeah, and now I'm just thinking about so many different aspects of our daily life. For example, you go to a website that's littered with trackers and JavaScript and all this stuff that loads up and pegs your CPU, right? Because it's so busy trying to load up all this garbage that maybe you're not blocking. multiply, I mean, if we're talking about something like forbes.com, which has a ton of them, which is a ton of garbage on that site,

you multiply that by, I don't know, 100 million people?

There was a art project by, I think her name was Joanna Moll. She did a analysis of how much data is transmitted when you just buy one product on Amazon.

Oh my god, this is gonna scare the crap out of me.

Joseph (31:58.214)
I mean, it's insane. I don't remember the numbers now because I didn't prepare to talk about this, but it reminded me when you're talking about this is that their business relies, that business model relies on getting really detailed information about consumers. And the tracking that's happening through JavaScript was in megabytes of information.

it in front of me. It is called the hidden life of an Amazon user. And I'll have a link to this in the show notes. So it tracks the purchase of a book. It doesn't matter which book. But in order to purchase the book, the Amazon website forces the customer to go through 12 different interfaces composed of large amounts of code. This code carries out all sorts of operations such as organizing and styling the site's content, allowing interactivity and recording.

the user's activity. Overall, I was able to track 1,307 different requests to all sorts of scripts and documents, totaling 8,700 A4 pages worth of printed code. The amount of energy needed to load each of the 12 different interfaces along with each one's endless fragments of code was approximately 30 watt hours. That's to purchase one book.

on Amazon. Of course, of course making it worse is all that code that that 8000 pages of A4 code is pretty much all used to like track your data and you know, make you the product as well.

And it's crazy to me that this has become the norm, that it's become so normal that this is the, what some might call the cost of business, And I think this is where trying to think about what we would accept in other domains of life, like would you accept this if you went to a store and bought a book and then found out that this is what was happening? Like, would you accept this? But we somehow, it happens online and we do.

Jason (33:56.886)
Yeah, yeah, I hear you. hear you. We're always talking about, we're talking about the benefits of Linux and FOSS, right? We're always extolling the virtues of, you know, of privacy, customization, freedom, right? But we are never, ever, ever talking about the power efficiency of a FOSS app versus a proprietary app. Why do you think that is? Okay, well, I know that we can't answer this.

in any way that resembles a comprehensive answer because we just don't have the data. But do you think there is a power efficiency advantage when it comes to FOSS software because it's not interested in serving you ads all the time and tracking your data and making you a product?

That alone, absolutely. mean, that is going to be consuming energy and it's energy consumption that's not necessary for the functionality of that software. Right. So, just the fact that that's turned off is probably already saving a ton of energy. One of the things that I focus on in the talks that I've been giving is the sort of inherent qualities of FOSS.

It's not always in terms of just efficiency, but also in terms of sustainability and that sort of there's many components that make something sustainable. And the ways that free and open source software gives users the ability, for example, just to switch software products if they want to because of data formats following open standards or interoperability.

as a value that tends to exist in the free and open source software compared to the proprietary software world allows users to make choices that they can't make if they're using proprietary data formats, for example. You're locked into that software whether it's efficient or not. So you've lost the ability to decide how you're computing, how much energy it consumes.

Joseph (36:07.086)
It may seem obvious, because with free and open source software, we've already thought about these things for decades now. But we haven't thought about it in terms of what does it mean if you're locked into a software product or a company or a government is locked into a software product that is inefficient, right?

And as you pointed out at your talk, then you start thinking about beyond the lack of energy efficiency, so that unwanted data on the cell phones in the EU, that data point that you were talking about, how much more quickly does the lifespan of a smartphone battery get reduced by that, right? How much faster do our PCs...

experience that both the drives and you know the RAM and everything in the CPU. How much faster do our PC components experience like wear and tear due to a lack of power efficiency? And the more I go down this rabbit hole, I'm like everything I've been doing in my daily life, I'm calling into question. I've stopped using my desktop PC. I'm not using it anymore. I have a Ryzen 3900X.

system with 64 gigs of RAM and a really powerful Radeon RX 6800 XT, right? Just a badass gaming PC basically and production PC. I've stopped using it. I was doing just little, little tiny tests and I actually have one for you. Check this

to hear more about this because you've been posting about it on mastodon and it's exciting. It's exciting that you're actually...

Jason (37:43.022)
haven't done a lot of testing yet, but I've done a little bit, okay? At 1 % CPU utilization on this PC, this tower PC that I have, 1 % CPU utilization, it consumes 67 watts. 67 watts, and like my Steam Deck at full bore, running something as powerful as it can, is using like 24, 25 watts, right?

Check this out, this is where it gets absolutely insane. So I ran the Dirt 5, it's a racing game, the Dirt 5 benchmark at 4K resolution on the highest preset and let the frames per second just go wild, right? It's no cap, just go as high as you can. 430 watts of power at the wall. And then all I did was I kept the same graphic settings and I reduced the resolution to 1440p.

and put a 60 frames per second cap and the power was reduced almost by half. By half. Now, after your talk, now I'm thinking, oh, that extra 205 watts is a big freaking deal. When we apply that out to like, in 2022, there were an estimated 1.8 billion PC gamers.

So now everything I'm doing I'm like, man, how much does my computer use when it's just in a sleep state? How much power can I save if I you know watch a Netflix stream at 1080p instead of 4k like it's just sending me into this Yeah, okay. I'm gonna calm down take a breath

I totally get what you're saying and I think it's great that you have started thinking about these things and that you're talking about it in a podcast so other people can hear this and start thinking about it. I would add though that we shouldn't individualize the responsibility to make these savings, these efficiency savings. mean, of course it's good to have individuals, but if we really want change, it needs to be at a system level.

Jason (39:54.998)
It is, no, you're right, you're right. This has been a lot of doom and gloom. we, let's turn it around into a little, let's get a little bit of positivity in the mix here and talk about KDE Eco that is, you know, promoting software sustainability. Well, the multiple organizations, KDE and yeah, yeah. And Blue Angel and others. Can you talk a little bit about the work that KDE has done on that front?

So we've really been focusing on orienting ourselves to the Blue Angel eco-label initially. Not only that, but that's the project that I was in was a government funded project from the German government, from the Federal Ministry for the Environment, Nature Conservation, Nuclear Safety and Consumer Protection.

And they funded us through the Blue Angel, the German Environment Agency, to focus on spreading and collecting information related to the Blue Angel eco-label. So a lot of the work has been oriented to that, which is a particular methodology for measuring desktop software, which KDE excels in. And looking at the

qualities that make for software to be more sustainable. So the KDEco really got started with this Blue Angel, Blaua Angle for FOSS, or BE for FOSS for short. But there's other sort of projects involved with that. There's the Free and Open Source Energy Efficiency Project, which is more generally looking at tooling for measuring software, for running scripts so that you can get reliable data, so that you can replicate.

the use that is driving the energy consumption so that other people can also do measurements and hopefully replicate the data. Those are the sort of two main projects that started the Kitty ECO. And the work that I've been doing is mostly, so there's in the community, you know, hundreds of people who are much more expert in software engineering and

Joseph (42:12.214)
all things related to this topic than I am. And I've been mostly trying to organize and support and enable work in this area. And that has been mostly involved in tooling for a lab and setting up a lab, which we have now set up and are currently supporting a student who is developing a remote access portal for that lab.

that's fantastic. So what what do you guys measure? Like what what is a typical session look like at that lab?

It's mostly in the preparation. So the lab, when you're actually in the lab, it's just running scripts and collecting the output from an external.

Okay, so you have scripts for K-Mail, right?

There are scripts for K-Mail, Ocular, and Krita, the natural painting program, that have been run and measured, and we have the reports. And they're available at the Free and Open Source Energy Efficiency Project repository.

Jason (43:13.846)
Is it realistic to expect anyone to take that data and compare it to proprietary alternatives like, I don't know, like Outlook, like Adobe Reader, which is probably an absolute resource hog compared to Ocular.

Absolutely. So the process is easily transferable to any software product. What will get measured using this particular process is the local PC running the scripts. You're not going to get information about data center processing. So it's not a system for measuring distributed systems. There is a methodology being developed right now by the Umfeld Campus Birkenfeld for measuring distributed systems.

but that's not right now what we're doing in the lab. So if they're outsourcing any processing to a data center, you'll only get information about network traffic that's also being collected in the process, but you won't get information about the energy consumption happening at the data center. But the process is easily transferable to any software product on any computer, since it basically relies on an emulation script, which there are tools for.

and an external power meter. There's really no more magic to it. So you can take any software, write a script to emulate the user behavior, run that 30 times, measure that using a power meter, and then compare that to another software product doing the same thing. Of course, that has to be prepared. That's the hardest part is preparing the scripts. If you want to have a reliable comparison, you would want to run them on the same computer.

and ideally in similar conditions. So simple things like running a script in the afternoon when the sun may be shining might result in the fans running more often than not. yeah.

Jason (45:15.32)
Wow, so it's got to be a really, really controlled environment.

If you really want to get good comparisons, you'd have to pay attention to this or at least be transparent about it, right? So at least someone can say, the data might differ because of these factors, like it's a hot room. It's going to influence how much energy the PC is consuming.

Yeah, so ideally a temperature controlled space would be the best.

ideally similar conditions when running this if you want to have a fair comparison.

Why has no one done that? I'm still baffled. I'm still just like, what is the word? It's gobsmacked. You know, I'm gobsmacked that this is not a dialogue that we see.

Joseph (45:59.406)
So the German Environment Agency published a report from the Umweltkampus Birkenfeld and the Öko-Institut, so the Eco-Ecological Institute in Berlin, doing this. They compared various software products. Now they did it for research purposes and didn't publish the names of the software products being tested. But there are some reports that have done exactly this and the results are really troubling. I mean, you really see huge differences.

Really?

The example that I used in the Linux App Summit talk is word processors.

undisclosed commercial word processor.

Exactly. Who knows which word processor that was, but it's the proprietary one.

Jason (46:46.253)
It does not rhyme with bird.

And we're talking about four times the energy consumption. And really when you look at the energy consumption over time, it even becomes even more troubling that you see when the software is no longer being asked to do anything from the usage scenario script, it's still doing things.

At this point, seems like we are in such a, you know, the infant stages of this, of the awareness of this situation, you know, about, well, about sustainable software, about the eco footprint of software through your various devices. As you mentioned earlier, it's good to, you know, on an individual basis,

to maybe cut those corners and pull back the power usage if you can. But I think it sounds like what we need right now is more advocacy, more people like you appealing to developers.

and creating a culture of sustainability in software.

Jason (47:50.99)
How do people get involved?

So I'm focused right now on, I think, getting data. I think we can talk about this in an abstract way. And that's important. But to make decisions, to really make data-driven decisions, we need to test and measure. When we talk about sustainability, it's not, in many cases, it's not talking about saving the planet. As someone pointed out recently,

It's really about reducing some of the harm or making things less worse, but it's still costing something. And to talk about relative harm, which decisions do we make and what is the harm and which is the least worse option? We need data.

The least worst option. geez.

Because all of them are going to, when we compute, at least given the fact that most of our energy is not being provided by carbon-free sources, it's costing something. It's costing something to the environment, and it's costing something in energy consumption and hardware production and all of these layers.

Jason (49:05.984)
It's getting worse now, I think, when we look at the increasing habit of things like soldering RAM to a motherboard on a laptop. Things like what I consider to be pretty egregious hardware requirements for running Windows 11, for example. We heard from tons of people, I heard from tons of people who had bought

new PCs or laptops in like 2017 or 2018 that they couldn't run Windows 11, which is ridiculous when you consider, yeah, I'm running elementary OS on a 2013 Mac. Right? And that's a whole nother level to all this, isn't it? It's the e-waste factor and it's, if your software, you know, if Ocular,

can run on hardware that's 15 years old and Adobe Reader can only run on hardware that's seven years old. Well then, obviously, obviously we know who wins that battle. Right? It's just like, you know, the more, the more developers that come together and the more companies and software that allow you to run their product on older hardware means less e-waste. It's that simple.

Absolutely, yes, and absolutely and this is where I think you know we need to start so Document what are the minimum system requirements? You know make that part of software? Documentation you know make it so it's clear that this software can run on older hardware and make that part of the the consumer side of things that you choose software because it is more Resource efficient because it allows you to keep your hardware in use

That's part of it for sure. I think we data. need to start getting data. need to start getting... I would say starting to do that will help push this field forward. And then of course data-driven decisions about how you're engineering your software, choosing to do things in ways that reduce energy consumption, even if the hardware can support the more resource-hungry choices, choosing...

Joseph (51:25.944)
for the less resource hungry one because right now, in many cases, you don't even have to think about it, right? The hardware is getting more and more powerful that we can just keep making the software less and less efficient.

I knew that we were headed in the wrong direction when maybe it was 2014. I did this PC build at Forbes and it was doing a mini ITX form factor. For those who don't know, it's just a mini ITX is just a smaller motherboard. know, like basically it was the size of a shoe box. It was a PC the size of a shoe box that could still run any game at 1080p maxed out.

Then I always thought at that point in the PC gaming life cycle, I always thought it was just gonna get better and better and better, right? And more efficient and more efficient and smaller and smaller because that is where you see laptops going. But then fast forward to like last year when Nvidia is like, oh yeah, you're gonna need a 900 watt power supply for this and it's a triple slot card.

And they're going to need to make special large cases for this. And you need this adapter to go from your, like, are you kidding me? How did we, where did we go wrong?

But let's try to stick with some positive things. I mean, I do think one of the benefits of the user-oriented aspects of free and open source software, I mean, we can do something about this, right? We can actually influence and we can shape the development of this software and we can start thinking about how we use our time and energy for making software more sustainable, more efficient.

Joseph (53:11.574)
And that's a really important factor when you think about how do you make things change is what can I realistically do? And we can do this. We have decades of experience of organizing thousands of contributors. This is not outside of our reach. That's incredible.

It is incredible. you mentioned that the first steps are would be gathering the data. Where can people take action? How can people contribute to that first step?

If you want to follow at least a initial step towards some sort of standard methodology, you can take a look at the methodology that we wrote about in our handbook, the KDEco handbook, which is oriented to the Blue Angel, and follow that method, and that gives at least a way to start comparing software, not necessarily for the sense of saying this is,

They win, but just so we can start to have some standards. The method that we've been focused on are external power meters, but that's not the only method. You can use software-based estimates of energy consumption. There are various tools. I'm not very familiar with them. I know one that's being developed by a startup here in Berlin called the Green Coding Berlin, and they've developed a nice tool.

that allows you to get all kinds of data. It's free and open source. You can check out what's called the Green Metrics tool. But I do think getting some standards and I think starting to think about how we can standardize the methodology, at least in some level, so that we can all have the same conversation.

Jason (54:55.126)
I mean, I think that, I think that, you know, what you're doing is the first important step is having the conversation with people, people like me, people like the attendees of LAS or FOSDEM or, know, everywhere else that you go. And I think that's really important. And, I just want to applaud you for that because, you know, someone who, like me, who is a tech enthusiast never thought about any of these angles until I heard you talk about it.

Yep, awesome. That's great. mean, this is what this is and your enthusiasm is infectious. Like it's great to see what you're doing online and mastered on the posts that you've been posting about the work that you've been doing, just trying to see what am I, you know, what am I consuming when I do this and that? Another thing and I, and I, I'm so the project, the Blau-Engel-Pofoss project ended in March and I'm no longer working for that particular project. So I'm saying this independently of my job.

Eco-certification is something that I think is seriously worth considering for free and open source software developers. As I try to point out in my talk is that you don't have to be free and open source software to get the Blue Angel eco-label, but we have a huge advantage given the inherent qualities of free and open source software, which are recognized in the criteria. And I think it would be nice right now, Ocular is the only eco-certified product.

And it would be wonderful if if Ecolabels can be a mechanism to incur an adoption in places like government and for companies that are interested in sustainability initiatives. And I mean, beyond the fact that I want to see free software everywhere, if we can get more software that supports interoperability and open standards for data formats, right.

how would that change the environment in which we're computing? Just the fact that you can use an open standard, which is one of the requirements of the Blue Angel, that your software product supports open standards, means that you have so much more options. And to see people locked into these systems, which are not necessarily efficient systems, right? mean, the product that rhymes with bird, assuming that's what it was.

Joseph (57:17.998)
it was being measured, right? That's four times the energy consumption. Now every office that's running that is consuming, you know, so much energy just because they can't choose to change. So I think eco labels can push certain markets at least in certain directions. And I think it could be something that would be beneficial to think about if you're a software developer. The process is clearly defined. The lab that we're setting up can be used

for data-driven decision making and software development, but also for eco-label certification. So hopefully that will be available to free software developers online. It is available today. You just need to organize with me or someone who's located in Berlin to go to the lab and do it, but hopefully it'll be available to everyone very

Yeah, because you mentioned someone's working on a someone's working on a remote access to that, right?

Yeah, exactly. Yeah. And we want to incorporate it into the CI CD pipeline so that when you make a commit, can then, if you want, run a defined script that will then give you some report back about how that changes your energy consumption.

Wow, that's a fantastic idea.

Joseph (58:37.356)
Yeah, and these are the kinds of things that we can start incorporating these kinds of metrics into our development process. We have to think about how there's embodied carbon in every software product. All the energy is consumed by the development process.

my god, I didn't even think about that either.

And these are the kinds of things that if we can start getting data on, it would actually be fascinating to think about what is the embodied carbon of this software product, of this release.

And the tools as they do in open source will get more efficient and more elegant and hopefully more widespread and more feature rich and all of that. well, there's a lot of resources here I think that people can tap into. Starting with I would say the KDE Eco Handbook, which I've skimmed through and it's excellent. So I'm gonna put links to that Joanna Moll article,

the reports that you mentioned, the green metrics tool, all of that stuff will be in the show notes for this episode. I would implore you guys, if you're interested, to talk about it. Talk about it with other people, share the data, learn the tools that are out there right now, and maybe talk to Joseph too. Is there anywhere that you'd like people to reach out to you?

Joseph (59:55.094)
So if they want to reach out to the larger KDE Eco community, there's the mailing list in the matrix room, both of which are linked to at the eco.kde.org website. If you have questions, someone there can answer it. And if they can't, I'm sure that they'll get a good conversation out of it. If they want to reach out to me, they're more than welcome to joseph at kde.org. So maybe also to add.

The KDE Eco is sort an umbrella term for these various projects. I mentioned the Blower Angle for FOSS. There's also the Free and Open Source Energy Efficiency Project. Now there's a Sustainable Software Goal at KDE. The Remote Access Portal is also part of this KDE Eco initiative. But KDE Eco is really much more, you know, the Free and Open Source Energy Efficiency Project is not the KDE Efficiency Project, right? We want this to be something that's part of free software.

This is not, it's under the KDE umbrella. KDE is fully behind it, but we want to get as many people, you know, having their own initiatives, collaborating with us, doing their own thing, any way that we can, you know, make this part of our culture so that we think of free and open source software in terms of privacy. We think about it in terms of transparency and we think about it in terms of sustainability.

Yeah, and all of that is is very much in the spirit of where I met you was the Linux app summit which is for people who don't know is is co-hosted by KDE and by gnome so You know to to quote-unquote competitors, you know joining forces to to make stuff better

And there's lots of great GNOME developers doing really great stuff. there's, you know, Philip Withnall has like an analysis of energy consumption related to their conferences and things that we want to also start looking at in KDE. You know, what is the cost of our in-person meetups and things like this. There's a great talk by Tobias Bernhardt, I think, who has the post-apocalyptic computing.

Joseph (01:02:03.662)
looking at what happens post-collapse computing, I think. And it's actually, that's intense talk.

So I hope that this conversation has not convinced everyone to recycle their computers and go live in a cabin in the woods, but maybe that wouldn't be such a terrible thing. But if you do, if you do that, if you go off the grid, keep in touch with people. It's good to have a little bit of social interaction. But yeah, I will keep you guys posted on, especially when I do more of this gaming testing, I think that's a big.

Contributor to my personal energy waste or at least it was I really do think I'm gonna switch exclusively to just gaming on the Steam Deck Not to mention like running, you know, it's gonna be it's gonna be 35 40 degrees here pretty soon Celsius. So I Don't know 90 to 105 Fahrenheit And no one wants to be running these big power hungry heat emitting beasts in the summer, right just

Grab a Steam Deck, grab a Power Efficient laptop. You don't have to play everything at 4K 120. I think we're gonna keep this conversation flowing and hopefully we can get other people in the mix. what I wanna do is I wanna throw it out to the community now to everyone listening. If you have questions about any of this, I want you to email me and I will make sure that those questions get answered by the right people and that we kind of do a, we come back in a certain amount of time and

and have those questions answered for you. So my email is in the show notes, but it is Linux for everyone at pm.me. That's a proton mail address. So Linux for everyone at pm.me. If you have a question, just send it and I'll route it to the right people. Joseph, thank you so, so much. And thank you for fighting the good fight, most of all.