@@Felix-ve9hs what do you mean? maintainers refusing requests? In that case, no money changed hands, so it's not a hand that feeds you. Users demanding changes? In that case, it is, I just don't see how is it any different from the first comment.
@@szaszm_ The other way around: Users of a project demanding stuff from a developer who spends their time creating something. It’s always ok to ask, but when you act as if the dev owes you something, that’s just wrong
Gotta go BDFL on the ones who do. Not enough nerds understand leverage, they want something you have so you can make demands of them. Demands not met? Ban them, they’re not worth your time. Either they change tune or you have one less problem in your life.
Can't disagree entirely, but I see it used a lot in the context of "unexpectedly departed", which IMO for at least anyone who isn't intimately involved in the ubuntu-rockchip project, would be an apt description (apt, heh). That's the context I was thinking when I wrote the title, but I definitely see how it could be interpreted in a negative light (not intended!).
@@Level2Jeff I see where you're coming from now. I hang around in spaces where people don't use that word so thanks for the heads up Like 0Lone I associate AWOL with dereliction of duty so it was confusing for a second between the title and content of your video
Yeah, AWOL is not a good term here.. the video itself is interesting and I tend to agree with the points made, but the title implies a very different stance than is presented.
I'm an embedded Linux engineer and I work with Rockchip SoCs. Rockchip software support is shit. The kernel they provide is years behind the mainline kernel and it's full of bugs. Not only do they make no effort to upstream support for their chips, they actively make it harder for those changes to get up streamed. Apparently there is some premium software offering, but if you email them about it they don't reply. How that complies with GPL I don't know. Basically avoid unless you're being paid to do it. If it's not worth Rockchips time, it's not worth yours
Well said. I did get an orangepi a long while back but poor support left me frustrated (I didn't know about Riek's work). I can tell you I won't be purchasing any more Rockchip hardware as long as the industry doesn't bother to finish their own software end of the product. Ridiculous they'd expect someone to do it for them for free.
Ultimately, if the hardware vendor doesn't want to deliver Open Source drivers to the Linux kernel, you're in for a major kicking. At best, you'll need to reverse engineer with no assistance... at worst you'll get legal threats. The number 1 advantage of Raspberry Pi is that they pay an Open Source vendor to produce Open Source drivers. It stuns me people buy other boards and expect some random person to spend their life supporting your stuff.
Random people supporting your stuff is the whole value proposition for most users. Yes, idealistically it’s the ability to modify things for your own needs but the more popular the product, the less people there are who do that.
Collabora has upstreamed an OpenGL driver for the RK3588, and I've been told that a Vulkan driver is coming in a couple of months. There's a big chance that you can run the RK3588 with a mainline kernel in 2025.
There has been one very toxic software company who charges most ARM SBC manufacturers to deliver software support in the form of a working OS with regular updates for a limited time(like a contract) and doesn't allow competition to grow up freely, they always try to destroy rep of anyone trying to copycat them or doing it for free like most people do with these non rpi boards. It's not an easy job to do as a pro bono work for the open source community and there's always some annoying users who purchase a certain SBC board, constantly demanding new features/updates like you were Microsoft.
My personal opinion is that the board manufacturers should take a more responsibility for the development of drivers for the boards they sell. Many times they advertise features for the boards that are never fully implemented because there isn't enough support from them or the chip makers. As a customer it can be very difficult to sort through which features a board ships with versus which features are technically possible. Many people are afraid to touch anything that is not a Raspberry Pi because they have been burned by this. If Rockchip makes it difficult then don't use their chips and stop passing the pain along to the end consumer.
Yeah, but if I hadn't done that I would be a retired factory worker or tractor mechanic instead of someone who sold a successful software company. I'm not sorry.
That can be the exact same thing if you reverse it. In fact much more so. I'm honestly surprised if you don't know anyone that has ie. Inherited business they don't care at all and have driven business down. This is very common and stress is common factor on all of these. On the opposite people that have good careers are virtually ones that discuss about their profession in their free time. I pretty much dare to say bluntry that you absolutely can't exel if you don't have interest in subject.
"We don't have to do the right thing by you", "Ok, I don't have to anything about this project anymore unless I feel like it. Right now, I don't feel like it.". No one should be mad. He's playing by the rules. It remains to be seen what happens down the road in a few years.
As a contributor to multiple OSS projects, good on him. Far too often big business abuses the kindness of the community. Quite frankly, the biggest issue with rockchip going back like 8 years has been software support and a reason why I never recommend their products.
Back when I first got into Linux, there was still a free software/ open source split. I think we need a bit more free software politics to push back against this sort of thing.
I genuinely hate how people demand things of open source developers. I’ve seen it happen too many times and it’s downright terrifying how many amazing developers get stretched to there limits just because people are abusive.
As someone that has contributed to opensource projects (in a minor way) and put out a bunch of "source available" WTFPL licensed code ... if you say that "X solves Y problem" and it doesn't work people are going to be upset and tbh a lot of the tech community in general doesn't excel in the social skills department. I have experienced this first hand, spent a bunch of time solving the problem (there wasn't one really) and then never got even a reply or thanks for making it easier for them.
And this is reason why most open source projects have no documentation. Your project might be the best thing since sliced bread but it's useless if only you know how to use it.
Folks are quick to call out “open source theft” but not quick to support the originator. We see it a lot in the 3d printing world. Companies profiting off the hard work of devs who barely make ends meet. Thanks for making this video. It’s sad to see this happen but I know all too well how they are feeling and don’t blame Josh at all.
I did a open source project for a certain 3d printable storage solution. The amount of comments on Reddit that said to support the developer and how much awards those gathered was a bit heartbreaking when nobody actually donated. Sure, it is a bit rough around the edges, but there are a dozen daily users...
@ folks are great about talking a good game in the 3 d printing world. They are quick to attack folks for perceived license violations, but actually supporting folks in ways that matter? Crickets. It’s disheartening some times. Sorry you have fallen victim to it.
Putting corporations at the same level than individuals in open source licensing is the Trojan horse that will eventually destroy a project. Change my mind!
That's a legal landmine. If you're building something of note it's better to relicense under GPL to prevent this stuff in general even from individuals
its more deeper than that, open source licenses work almost like a peer to peer model in which eventually you get leechers that only take and don't contribute. Free software licenses are even worse, they put the user needs before developer needs, leaving everyone unpaid and scrapping by. Yes, I know those licenses require code to be contributed back, but code goes in, it cost time, resources and money, nothing comes out. Open source ends up being donation of time and money, hopefully you get code contributed back, but its basically a give away. That's a terrible business model, doing charity development of software. So, no, even if you changed the license so only individuals could use the software, there's still this problem that the developer never get anything back for their work, except the satisfaction of making a good software (because that's almost impossible in commercial settings) Maybe we need a new license model for open source software that's not a "free license" of use, something that require users to pay back in some crypto-currency for the use if its not a personal use. Nothing is free on this world, so why open source software should be free to use ? I want a free-license power generator, and chip foundry, and sand mining, if everything was free-license and resources were free to take from the ground, then maybe it might work.
After I said that, I read the Drupal , that's neat way of solving the problem, track the leechers and make it very public that they don't contribute. Maybe shaming works, and incentivizing the good contributors is a good idea, hopefully the pool of resources donated to the project goes to those who are maker and not to the takers. I still would prefer that sort of thing in the license (somehow), not as policy of a ONG that controls the project.
Knowing how engaged Joshua was in the community it just saddened me to see him feel so bad after all the good he’s done. I mean I have a Rock5B running rn with his image. He really is someone like no other taking up a task as big as he did and I hope he can find something that’ll give him the same spark initially
Technically the dev did not go AWOL because they announced they are leaving.. It would have been going AWOL if there was no notices of it or anything of the sort
AWOL can just mean 'deserted' or 'absent', though it often has the connotation of being unannounced. I generally hear it as "being gone abruptly" though, so that's the sense I'm using for this title.
Technically the term originates from not being given permission to leave. If someone goes AWOL, it might have been unannounced or they could have said they intended to leave. The problem is that they do end up being absent, and in terms of military service, or a job, it's not going to be a good time when they get back.
@@eirinym Indeed the question here is who must the dev seek permission so they have leave? I'd say AWOL could apply but really shouldn't as the dev you can argue works for the community, so they don't get to announce they are taking a break without community permission perhaps. But really I'd say the Dev has the ability to give themselves as much leave as they want - its not like they are working 9-5 for a regular paycheck being an open source developer.
It's absent without official leave. A military term. In short being absent without permission. Since this developer does not need permission to stop he cannot be awol. But it's fine if you take the more generalized meaning. But even in that case the announcement, makes using the term stretching it.
Interesting to hear about your experiences and setups. Right now, I am just barely skimming a burnout, mainly because I am the only developer in my company and have to develop a fullstack app because "the boss said so"... And, without me really knowing, went and advertised that fact. So now I am literally crunching to meet early december as a deadline...kool. Thanks for your insights! I plan to be more active in OSS and FOSS one day and write some fun software I need myself or could be useful to others - so, hearing how long-time-doers do it, is very insightful. ^^
Oh boy, that's a tale as old as time :O The crazy thing is, I've had some of those situations turn into a nuke that destroyed a project entirely... and others turn into opportunities that went on to be amazing for the startups-it's probably more often the former, though. Especially when you're solo, I think a lot of management doesn't realize just how low the 'bus factor' is. If you walked away today, would they still be able to launch? If not... that's a massive business risk for them, and they pay a lot for insurance to mitigate fire, or theft, etc. Why can't they invest in software? :)
It's why they invented "quiet quitting". Do what they pay you to do but no more. I'd be looking for another job at the same time. Work your 8 hours then go home. That's how you avoid burn out. Tell them early and often that its more work than one person can do. Do it in emails so it's documented. Then when it goes wrong you have have a paper trail. The reward for hard work..is more work. If you go the extra mile, that becomes the new normal.
@@eat.a.dick.googleOnly because you've never been on the receiving end of a spam PR, not to mention ones that just rewrite everything and demand you to merge it in as is.
Rockchip is a good example of the lack of software support. I bought some Raspberry Pi alternatives with the RK chips that are gathering dust at the moment. Nice chips, poor support.
Every time I see a video like this reminds me why I avoid ARM computers and I stick with x86 as much as I can: - no standardization whatsoever, I have 15+ years old x86 devices, long abandoned by their manufacturers (or having their manufacturers gone completely), that can still boot the latest vanilla Ubuntu or Debian (even Windows 11!!!), good luck doing the same on ARM unless some burned out unpaid volunteer decides to do something in their spare time. - proprietary boot-loaders and kernels that get abandoned as soon as the device is out the door. Does every ARM computer really needs it’s own proprietary bootloader, top secret encrypted device tree and so on just to boot? They can’t even be bothered to submit patches upstream to the Linux kernel or even to Armbian. Make these devices be usable longer? Oh the horror!!! We can’t have that! The only way to even use these things is to use the manufacturer’s bespoke OS, which they never bother to upgrade at all and only works on specific versions of their computers. Good luck even running the manufacturer’s software on different generations of the same hardware. Can’t they get their stuff together and figure out on a standard way to handle booting, hardware detection and resources allocation? This is a problem that has been solved for over 50 years for crying out loud. People like to hate Microsoft (with good reason for all the shady things they’ve done), but for all their failings, in the 90s and 2000s they forced some standards on all x86 manufacturers and now everything simply works (most of the time anyway). Maybe the ARM ecosystem really needs someone like that to strong arm them into compliance. ARM people like to talk about how ACPI, UEFI, BIOS and other x86 related things are bad and ARM is so flexible and superior. Sure, x86 and it’s associated technologies are a big pile of ... antiquated stuff that really need a thorough cleanup, but they are STANDARD for every x86 device, from the lowly industrial board hidden in some old dusty CNC machine to the fastest supercomputers ever built. I had some hopes the Risc V people, having the chance to start from scratch, would avoid these issues, but after watching the latest development in this field, I’m not holding my breath...☹
15 year old devices are a real energy hog. You could buy a Pi (other SBC are available with less support) with more CPU power for $55, use a 10th of the electricity, and get more performance as well. There is a law of diminishing returns on keeping older kit running. And of course, the Linux kernel keeps dropping features it thinks no-one uses anymore, so even those old machines will stop working soon. Its worth doing the energy calculation.
What’s interesting is on the “server” side there *are* standards, which is why we can boot the same base distros on Ampere/Graviton/Grace. It’s the desktop and mobile/embedded SoCs that are a mess.
This video was eye-opening to me. I was confused on how to do open source properly. But the drupal example really showed me how to fix the maker-taker problem. And what open source projects I should look out for, because I want projects that protect themselves from the burnout problem you describe At the very least I will start donating money to the FOSS I use day-to-day once I get a job next year
And also see if any of the projects you use have in-person meetups or conferences. I try to attend some every year, it helps to keep that community connection alive!
Hello Jeff. During my career I helped with many Open Source projects and at one time was a contributor to Drupal. What I call my greatest accomplishment was forking Cygwin to create MSYS which someone else forked to create MSYS2. It's easy to burnout if all you do is Open Source so adding other interests helps. If you feel burnout, back away for a few days then come back to it.
As a biologist I must say that parasites are part of the ecosystem and we must accept that... that said, it is also part of the ecosystem to build the best possible protections against them.
I've tried to shift my thinking away from burnout or success/failure and more just letting my interest in things run its course. It's not that I failed at some project I undertook, I took the project to the extent I was interested in, at least for now. Allowing myself to emotionally distance from projects and then come back later is very freeing. This doesn't help at work much, but my advice is to talk to your manger if you're experiencing burnout. They might be able to help!
Companies need to be prepared to fork the project in accordance with licenses or write their own code if they make the devs mad. If enough developers exert their control over the software (by not continuing maintenance) then perhaps the companies will analyze the risk of losing the developers and find ways to support the project.
Absolutely. A maintainer’s quality of life should always trump outside demands. Sadly too few adhere to that to try and do right by their users and while noble, it’s only worth it to do if you directly profit in cash. Exposure kills!
I've dealt with burnout in the past at a prior job and to a lesser extent at my current one. What made the real difference was being in a role that I wanted to be in, with potential for both personal and professional growth. To prevent it, being surrounded by passionate, caring coworkers who prioritized mental health, pushed for reasonable deadlines and workload, and actively encouraged us to say "no" to additional projects and duties without discussing amongst the team first. I recovered, and due to my team's culture, we prevent it from taking hold in anyone on our team. The community we surround ourselves with absolutely makes or breaks burnout, and to see such a great contributor experience it due to unreasonable demands is such a loss. Thank you for bringing attention to this issue, Jeff. It's truly crucial.
Jeff, Truly a sage take on the whole industry at this moment. The sales, marketing, manufacturing, megastructure of capitalism fails to feed its golden goose - it's a dynamic I have seen in the industry for as long as I have been participating (Since `96). How to pay for development? How to balance the demands of business and production against the creative industry of coding. In hospitals they routinely do not pay healthcare providers for time spent writing patient notes and communicating with other providers - too detailed a note takes too much time to write and read from the perspective of the business. Fundamentally writing code takes thought and time, and business still doesn't want to take "longer than necessary" to do that task. We all have heard the "Ship It" mantra. Where what makes for good code and a good tool is thought, refinement, followed by more thought, and more refinement. Business wants "something to sell" and at some point doesn't really care what it is. That's not healthy. If the business apparatus can't see fit to work with its builders and nurturers, then strong boundaries, and stepping away is the only solution.
Supporting software is very difficult, one thing I learned from several decades of doing it. BTW I'd love to see more about running Linux other than Debian on ARM , especially Rocky Linux. Much better than the Debian side IMHO
When you use open source you absolutely are not entitled to any support. The audacity of people making feature requests, there shouldn't even be request trackers for open source software. You want something fixed, you do a pull request.
@@superchiaki 100%. I was just thinking that, there was a fedora that could run on a Pi, so there is a template to go by. I am now a retired software engineer and I spent a lot of the last couple decades building custom Operating System distributions and porting systems from one OS/architecture to another. I've upstreamed open source before to several projects, I have considered throwing in my hat do do more.
@@monad_tcp Well, if you do a pull you should push it back up. Just whining is not helpful, but many don't have the skills to work on software. They should not be cut off as long as they are giving useful feedback on problems and feature requests. Maybe if you can't support the code you should kick in a little bit of money.
This video proves that you've got good ethos in your work but also that you're a excellent person all around. Good on you for speaking out about this. Open source projects need support, collaboration and kindness, not nastiness and gatekeeping. I agree with Veronica (Explains): Linux is awesome, and so are you.
To avoid burnout is to avoid falling into voluntary slavery. In tech that means scheduling a fixed 1x24hrs day off every week and having something else todo that has large non-tech content such as charity.
It's getting harder and harder with security patches now considered as important issue. In the past, volunteer developers can work at their own pace. Nowadays, it's becoming dangerous not keeping the software updated.
@@bltzcstrnx If it's that important, then they'll provide support for it. There's a lot of times where people will scream about urgency of something but if they're asked to help, suddenly the urgency fades away. Unless you're paid to be on-call 24/7, don't be.
I recall there's a cdn provider (that's open source) that provide package distribution service that people use on their website, like jQuery, well, the cdn went down, the moment the creator went to sleep, and the cdn breaks tons of website, it went down for 13 hours iirc, so basically that xkcd comic lol
With regards to avoiding burnout in my career as a computer engineer before I switched fields due to losing the love for doing it 9-5, was by scheduling my time in a calendar and sticking to the time I alotted to various things I needed to get done, and wanted to get done. This allowed me to stay on track and make sure everything got the time it deserved and when it was time to walk away at night and wind down and spend time with the family it was considered sacred and was scheduled. No one is useful when they aren't at their best and burning yourself out only works against you. We aren't machines so we can't try to keep up with working like them. Finally, in my case I ended up Paraplegic at 42 due to an unfortunate accident, and the first thing I thought was, thank goodness I took time to spend on doing things I wanted to do that I can't do any longer because of my impairments and pain levels. My life changed forever in the blink of an eye at 42, and we need to remember something like that could happen to anyone and it really puts things into perspective. At least hindsight was 20/20 for me after it happened to me. Cheers!
Love the Cardinal Glennon shirt Jeff! You forgot to mention your health issues you have faced over the pasted few years that have also added to your plate when building your base .
Our circles overlapped years ago when I managed drupal for Canonical. I shifted my focus and became an open source lawyer. If you ever want to do a collab and talk about open source and legal stuff, I’d love connect. We share a similar philosophy on this. And I too have seen and felt this burnout issue.
In February of this year I was quickly, ludicrously quickly, heading for burnout at my IT job. Most of the burnout was being caused by a certain person from a team I left due to toxic, toxic behavior from them. I'd finally had enough and resigned my position and took many months off from work. Only way I could survive.
If you care about money, don't do open source. It's simple as that, when you start making open source projects you are aware someone might take it and make millions off of it and give you nothing. If you can't deal with that, don't make the project open source.
I used Citra a bit, but mostly gave in support time and effort And then the lawsuit happened, but I am happy I helped propel that project on, even if for a bit
I've made a few open-source projects that have become popular, and I agree wholeheartedly with your video, especially around community. I think as well things that are used by the end-user are a lot more challenging, because your community becomes largely comprised of people who aren't necessarily technical. It almost becomes a chicken-and-egg problem where you want to encourage contribution by making your app better, but at the same time you don't have the resources (especially around UI, testing and localisation) to accomplish that. OSS contribution doesn't have to just be about contributing code - I am very thankful to the graphic designers, translators, testers, and community managers who donate their time to ensure developers can keep developing.
It's great hearing your view on its all. Burn out is real. I like to establish a philosophy around each new project I start; and commit myself to it it. Otherwise if something gets popular, it's easy to begin to resent the lack of financial gain or contribution. If you're doing it because you like building things people use, then remind yourself why that's why you're doing it, and be prepared to take a break from the project if it gets too much. Jumping between projects and personal hobbies, even if everything as a whole slows down, has helped me to continue enjoy working on it all. People "stealing" code or trying to compete with your project, using your own work, can be frustrating to some, but I've found that persistence and maintaining the open-source mentally will protect you. Sometimes you can find more strategic ways to protect yourself here, but most times these stresses come and go without real detrimental outcomes. Surviving financially as an open-source contributor can be hard also; balancing your personal needs isn't easy, and I can understand why some people might start to resent the situation if they feel they've sacrificed so much without improving their own personal outcome. I feel very fortunate to have the privilege to work on open-source, and not need to stress about finding ways for it to make me money, as I can get by with side hustles and being content living a simple life. I've long given up on some dreams, and accepting things as is will lets me enjoy what I already have more. GitHub sponsorships and other gifts pay for about 10% of my bills, which helps, but it's still not sufficient alone. It's definitely not for everyone as a result; not without some UBI system at least, and that has its own implications I'm sure.
Very tough subject. My personal approach is to document as best as possible what I’m doing. This way I get way less issues, and new contributors feel empowered taking it over. So I don’t feel too bad stopping working on one of my OSS projects or even giving away the bdfn ("now") title. And as you perfectly explain, ignoring is sometimes the best way to keep your mental health.
Thanks for talking about this. It has been a good 20 years since I've "taken point" on a large FOSS project, and damn do I completely lack any ability to miss it. That said, there are things one can do to help themselves that should get top billing: • When somebody is willing to contribute, even if it doesn't fit exactly into the ideal process, strongly consider taking them up on it. EG Many people who can contribute a bugfix patch don't know the first thing about writing tests. Sometimes a clear and complete user story is golden-reserving the community for only those who can write code and tests and documentation is asinine. • Most people don't want to you to fail, they just suck at communicating it. Keep this in mind and it is easier (for many) to not take things personally. • Many of the best contributions come from people who don't author anything directly, they engage with others to help them make better stuff instead because that's what they're good at.
The flip side can also be a challenge: it's often hard to donate money to FOSS developers, setting that up is just one more challenge. My current job is shit, be in the past I've had an annual budget to donate to FOSS we rely on. These days it's just my money. My Ask: please, set up an account somewhere, somehow, that people can send money to. IBAN, PayID, even payal if you have to. It might only be $10 a year, so don't put lots of work in, but put it there just in case.
I think the main cause of open-source rugpulls, burnout, etc. is that there's not really any straight-forward ways to license source-available software that isn't open-source too. If people expect more back from the work they do than the license requires (and then don't get it), that's going to lead to burnout. Or people only can afford to hire a lawyer at a certain size to then write a license that actually expresses their intentions. Having open-source as the only way to license source-available software isn't good for anyone involved, including and especially open-source advocates.
Yeah we have hundreds of SBC now, and so much of them are utter trash when it comes to OS support. I wish companies would quit dropping 10 models a year with piss poor support (Orange, Banana and Friendly are notorious for this). Its also something thats often missed during reviews of new product because the reviews are so shallow. They create a baseline build of an OS and then just abandon it. I am not a PI Foundation fan because of the way they have acted over the last decade but at least they have enough support to have a continuous release of updated OS's. I'm sorry to hear this dev got burned out. With 31 years as an IT professional, its a sad norm that most of us will experience.
Interested to understand the phrase " because of the way they have acted over the last decade". The only thing that comes to mind is keeping business alive over covid/silicon supply crisis by prioritising industrial customers (and by that I mean any industrial customers, not just those buying thousands of devices)
@@jamesh9756 When I first started using Pi foundation products their focus was ostensibly education and hobbyist. They gained a lot from the community that sprung up around their products, in part contributions, code support, projects built to take advantage of their platform that also drove a lot of interest, etc. all while maintaining their low manufacturing numbers that led to constant shortages on launches and for a lot of each product life cycles. Some of us had great ideas, and wanted a way to purchase bulk and implement and sell completed devices and projects on a small business scale. When we tried approaching about increasing purchase limits even in novel ways (like having to pay for matching units for charity) they said no. They said their focus was on the individual. So we lived with that. They still grew in popularity in large part due to the community. Then they made the call to focus on industrial and commercial customers. As someone who has contributed to OSS at times for a company to be all, nooo really we are about the individual... so individuals contribute while they grow, and their product and reputation gets enhanced..... and then give people the finger and say, nahh just kidding mate, we all about the industrial customers, but thanks for what you did to get us here. Its as much their handling of their customer base, as the proliferation of cheap MCU that is why there are so many competitors now. So yeah its about the attitude perceived about a business. Like it or not, I see this as them saying Let them Eat Cake. So I avoid their products now. I will gladly burn 100 hours doing compiles and making a competitors board suit my needs. My opinion. Not looking to argue, just how I feel after my dealings with them over the years.
This might be my favourite video of yours. It's very candid, and honest. I also ran across you a few times when I was also involved with Drupal, and agree with your thoughts. Burn out is real, the maintainer for ubuntu rockchip was very kind, so was really sucky seeing him suffer.
brb, I'm gonna go sell your ansible roles /s (btw Thanks for those. That's how I found you, actually. Though I thought you were older when I originally found your ansible roles lol.) edit: Also I still don't really know what Drupal's good for personally. It's never made a list of ideas or platforms for projects (partially, likely because I don't like to do PHP that much.)
Drupal used to be great for blogs, small mom-n-pop websites, even small catalog sites and medium business. They moved upmarket 5-10 years ago and now it's mostly for custom enterprise CMSes. A lot of big sites still run on Drupal, and my blog does too... but my blog is a horrible use case for it (IMO).
It's similar to Wordpress, lets you quickly spin up rich dynamic websites, and has premade apps for common use cases. And it can be customized in PHP when you need to, without needing to write the whole website yourself. All of my experience with it is customizing and hosting a simple Drupal website 8 years ago.
That post (along with everything I know about ARM SBC’s) is why I will only buy RPis for now. You can make banging hardware, but it’s a paperweight without support. It’s why Apple laptops could get away with 8 GB of RAM for so long (literally until they needed to run AI locally)
Changing license is fine as long as it's done legally. The old versions that were released under the more permissive license can still be used and changed under that license, only the newer changes are more restricted. MIT -> GPL or MIT -> proprietary/source available/restricted can be done without too much effort. The other way around requires a new license from all authors / copyright holders. But either way, it's usually causing community outrage, because people don't like when they used to get something for free, and they no longer get the thing for free. (Myself included, moved over to OpenSearch and Rocky Linux.) My software philosophy is similar for my own code that I write as a hobby: If I'm giving it away, I'm doing so with no strings attached. If you can make more money from it than I could, good for you. I don't care about attribution either. So it's usually public domain, Unlicense or similar.
Changing from FLOSS to proprietary is bait and switch. Why don't just make it source-available or proprietary since the beginning. Unreal Engine do that, and they have many contributors. Personally, I don't have problem with developer choosing their own license. What I don't like is the recent bait and switch. They use open-source for clout in the beginning, then change the license when it got big.
The distinction on my side is mostly on production effort and less topic... I figure the YT algorithm decides what it wants regardless of where I post the video these days :(
Rockchips biggest problem is its inability or willingness to support its own chips. On paper the chips are great. But they simply do not provide even reference level drivers for the hardware they make. How do they expect people to use there chips on mass scale? Now that the key open source devs are walking away because their work is ether stolen or not appreciated. .. Rockchips going to have a sobering moment I think. Also the whole makertaker situation is one of the reasons I'm reluctant to release my code. Always wary some one taking something I've put my heart and soul into being taken by some one else and making money off it.
You should hear Brodie's (with his Tech Over Tea podcast) interview with Michael from Futo. He sounds about as ideological as RMS, but focused on making sure Free Software developers have a reliable source of income, and that seems to be the raison d'être of Futo, which includes exploring alternative licensing that can still give users all the same FSF freedoms, but can help guarantee funding.
This is a old war, because of licenses the Unix have been "forked" and now we got Linux... the future will be divided into full open-source/hardware or full locked systems 😮
In case folks don't know, "🚩 The Anti-Capitalist Software License" is a thing! Not sure it would help in this particular scenario, but... it's a thing to be aware of for those putting new projects into the world, and/or considering changing licensing for their existing projects.
I don't. In the very early 2000s, I wrote a bunch of perl libraries for music tagging that were useful in my personal project. It was a huge mistake. Yes, lots of people used them, to my surprise, but I didn't have the temperament, the time, or the social skills to maintain these libraries. Creating an open source project is, I discovered, a way to turn your hobby into a job, and when you have pathological demand avoidance, a job sucks.
Why is support on arm so spotty when anything x86 just seem to work from the get go. Does Intel and AMD do the work or are arm chips so different each one needs work.
X86 is much more standardized with ACPI and all that. Very little effort outside of drivers needs to be put in for it all to just 'work'. Arm devices are all slightly different and custom so a lot of the time you need to do a lot more than just drivers to get it to even boot Linux.
when arm v6 came, everything had to be rebuild, same for v7, and nothing is that much diffeent now every company does their own arm stuff, every couple generations everything breaks and has to be rebuild for the new chip x86 is big and heavy because they dont do that, they keep the old modules, the old stuff and add on top, this is why arm can be ligther and power efficient, they get rid of the old, and invent again, but forget to offer drivers on x86 alot of 16 bit stuff still works, up to 64 bit, all is there if the os can handle it
x86-based systems are almost always PCs and PCs have been historically standarized when it comes to their pre-boot environments: UEFI (BIOS in the past, though many UEFI implementations still offer 'legacy BIOS', so it is possible to boot old, good MS-DOS on modern PCs), ACPI etc. That is, standard interfaces exist to provide OS loader/booting OS kernel with basic HID, I/O, video and means to obtain info on available hardware. Therefore, on x86 generic OS images can be used, because no prior knowledge of hardware is needed to blast off. x86/PC is rather unique with this luxury of freedom of choice. ARM and RISC-V worlds are much more locked-down and in the segment of SBCs practically devoid of standarized preboot environments.
This is good advice for all professions! I'm not good about it. Lots of folks need help, but it's not our responsibility to help everyone. That's just reality. I could, at the expense of my family, health, and sanity, do pro bono work 40 hours a week (on top of income-generating work) but that's just not realistic or healthy.
This does not go nearly far enough. Desktop Linux will never become a thing without consolidation, and thus also bundling of forces. Because confusing and overwhelming the customer is poison for the success of the product. - As the saying goes: “If You Confuse, You'll Lose”
I’m big into the DIY eurorack community and the owner/proprietor/sole-employee of Mutable Instruments, the biggest inspiration to the entire eurorack community, completely stepped away from the game. She made her designs completely open and a year or so before she closed up shop she was invited by Arturia to corporate gathering and 3 months later Arturia released a product with her code and used her brand in marketing with zero compensation. Her products were then given the same treatment by Behringer without even getting the free lunch. I get that the license doesn’t require compensation but then there’s also just being an asshole. Then we don’t even get whatever these leeches do with our improve things because they won’t open source their product. I think open source is the reason we have nice things but when that ecosystem can’t actually be fostered I genuinely wonder if we’ll continue to have nice things.
Korg has been caught using the Linux kernel without providing their changes, as their OASYS, Kronos and Nautilus keyboards are Intel Atom PCs (which can be upgraded like PCs, since it's a bog standard Industrial ASRock motherboard inside the case. Yamaha does use MontaVista Linux in some of their hardware and you can request the source code if you want.
Now that our house construction is finally finished, I finally got around to trying out the Orange Pi 5 Plus 32GB that I bought months ago. And what did I have to realize? The hardware is great, but the official software support is total crap. I am extremely grateful to Joshua for his work. Without his valuable work and time invested, the hardware is almost useless. I completely understand his reaction. It's a bitter loss and I really hope he comes back. But your own health comes first. You do something in your free time with extreme dedication because you have fun doing it. At some point, those around you see this devotion as normal and begin to make demands. The tragic thing about it all is that this is not limited to IT or open source topics, but happens everywhere. I experience exactly the same thing in my IT job, but also in my private life. At some point it just doesn't work anymore and you need a break or stop doing what you actually love completely because otherwise you'll just break down. Get well soon Joshua!
Yes I agree, open source software has given us a lot of nice things and opened ways for people to widely benefit off of each others work. But the sense of entitlement that we as open source users can have can be very bad.
"...and if you don't like the way that I do it, fork it!" I like what you did there. 😂 Also, I really like the liberal with your no's bit. I'll have to try that. Thanks!
I wrote a telescope driver for the Meade brand of scopes. I did it as I had one myself and wasn’t happy with what was available. 5 years later, I’ve found myself in the position that I don’t want to use that scope anymore, have an different one that works better for what I like to do. But, I’m now stuck supporting this code that I wrote. It feels more and more of a burden, but I don’t want to abandon the project until I don’t have access to the scope anymore.
I think you should take @dareelcatskull 's advice above (in a comment) - don't think of it as failing, but letting your interest in that part of the project run its course. Make all the artifacts you use to create that driver (no matter what state they are in - no don't clean them up!) available, and step back. The only one trapping you into supporting that code is you.
The more software ecosystem issues I encounter, the more important standards seems to be, having communities build with the same compatibility requirements seem to be the only alternative to high resource businesses
Open source foundations are a good idea. Basically they do the work of canvasing for donations and support from companies and then redistribute that money to projects that need it. Specifically to avoid the XKCD situation where a single developer is propping up a huge amount of downstream development.
The only reason his work is awesome is support for hardware acceleration for jellyfin/immich. Crazy how RK3588 spec sheet of encoding/decoding reflects the actual use (FINALLY) using his stuff. It makes it an option compared to intel N100 for jellyfin specifically.
What is not fair is when they take open projects add to it and then close it so no one else can make from it,, keeping all the spoils for themselves. Talking about burn out do you ever see your family? You run two channels, do projects, write software and up load to github and build a studio workshop.
One solution to this is dual licensing. As copyright holder, you allow anyone to fork your code and use it without restriction as long as changes are fed back to the community (AGPL rather than GPL, so people can't avoid releasing the source code by keeping it only on their servers and not technically 'distributing; it) and a second commercial license for people who don't want to contribute their changes back. The problem with dual licensing is that you need to either have a single developer, who re-implements any external pull requests, or have a contributor license agreement in place, which assigns rights to the copyright holder. In either case, it's best to be up front about dual licensing at the start of the project, otherwise people will feel cheated if yo try to introduce it later on.
Compartmentalization helps me there - it was a lot harder before I had a separate studio to do the RUclips stuff! It's always tough to keep away from the allure of 'just working on that one little issue in that one little project' (which always turns into like 3 hours debugging lol), though.
i think missed a major factor in Joshua's post... money. he clearly states that hosting the project is expensive and he can't afford it. community and not getting burned out are important, but i am seeing a lot of FOSS people with major projects leave the community as they simply can't make ends meet while working 90 hours a week...
30+ years in professional software development here. I got burnt out on Project work, the deadlines, the testing, the go lives and attendant stress, warranty fixes then rolling into another. My boss thankfully noticed before I had the balls to admit it. I was moved into a senior support role which I love and have been doing for 4 years now. Lesson, I guess, is to speak up if you're struggling and let someone else help out. I was considering leaving IT completely, glad I had a good leader.
You are fortunate to have a boss who cares. I think that is the exception, and not the rule, especially when the head of the company is ruthless... it filters down.. sometimes slowly, but eventually people who last adopt the mindset/approach.
I have managed several open source projects and I just handed them all off one day, or archived them because I just couldn't do it anymore. People demanding things, being super pushy and just not willing to put in their own time to add commits but just whiny little asses. Yeah no thanks.
I know exactly how this dude feels, some of us do things for the wider community for free, but never receive the respect or thanks we sometimes deserve, some companies leverage their business models on things found in the opensource community but never give anything back, whenever I see someone fencing some new SOC or SBC on RUclips, I always ask them if the manufacturer has posted the source code to their operating system, have they provided the source code for the driver stacks, the answer is either silence, or its no, we need to stop helping these ungrateful companies and only work on hardware that does provide everything in the open and also assists and gives back to the community.
I get what you mean when you say "you're happy if people can do something awesome with your code" but your followers having to "chime in" (pay you/ donate to you) while these companies making money using your code aren't reimbursing you, means your supporters are essentially subsidizing corporate revenue. If these companies would simply pay you (even a small fee) for your work, your followers could either safe their money and still get awesome content from you or their money could be used to increase project budgets or production quality. Either way, they are directly exploit your "free work" and therefore by extend, also the people supporting you.
Honestly I feel like we need more licenses where it's like "this is free to use and modify, but if you're a business and your company makes over $1,000,000 a year you have to pay royalties" or something along those lines. Open source and free software (both versions of the word free) are extremely important for society, but developers also need to be compensated fairly for their time and effort. So those who profit off their work should share some of those profits.
In other news I probably need to start contributing to open source projects myself. I'm a recent computer science graduate and I can't find work, so now's a good time for that (though I'm also still just recovering from the burnout my education caused). No clue how to even get started though and not sure I'm ready.
I wonder if a kickstarter could be made by the original developer for "revisions" Want a newer version...pay $$$ to support the developer...once the target is reached, the revision is released It's time to "buy them a coffee" to help support those intelligent and ingenious enough to release code that makes all our lives easier
I'd rather send money direct. Running a kickstarter is a lot of work and risks the developer getting even mroe burned out. It's also a different skillset and adding that as another barrier seems like a bad idea. What each of us can do is contribute money if we can. It makes a difference, both emotionally "someone cares, also someone likes what I'm doing enough to send money", and the obvious "I have money" way.
Rockchip wants to live in their comfort zone. They don't care about new products because those don't move that many chips. Their only metric is chips sold. No future prospects, no insight. The same exact thing Microchip did. Their history is Rockchip's future. Sent from a Fydetab Duo running BredOS. This stuff is dope.
I like to get myself a very comfortable life with the stuff I build. Why not other people as well, if they help me with my stuff? Sadly, this really isn't the case for all open source communities. It's really hard to build nowadays. There is so much competition as well. So much VC money nowadays.
We had similiar stress and loss of motivation with LZMA SDK as far as I remember. This was purposeful work to hijack the project and the world depending on this SDK. Do we have similiar simptoms here ?
Making demands of an open source developer is abusive behaviour. I’m very irritated on his behalf.
How about them (the users) not biting the hand that feeds (the developer) you?
@@Felix-ve9hs what do you mean? maintainers refusing requests? In that case, no money changed hands, so it's not a hand that feeds you. Users demanding changes? In that case, it is, I just don't see how is it any different from the first comment.
@@szaszm_ The other way around: Users of a project demanding stuff from a developer who spends their time creating something.
It’s always ok to ask, but when you act as if the dev owes you something, that’s just wrong
@@Felix-ve9hs Not sure everyone always should ask for everything they want, all the time.
Gotta go BDFL on the ones who do. Not enough nerds understand leverage, they want something you have so you can make demands of them.
Demands not met? Ban them, they’re not worth your time. Either they change tune or you have one less problem in your life.
AWOL is a bad term to apply here, I expect. That implies a dereliction of duty. A hiatus of a volunteer is quite a different matter.
Yeah, quitting a volunteer job with notice isn't going AWOL. It's simply deciding not to do it anymore.
The military hasn't used AWOL in decades it's UA. (unauthorized absence) but yeah the sentiment is correct.
Can't disagree entirely, but I see it used a lot in the context of "unexpectedly departed", which IMO for at least anyone who isn't intimately involved in the ubuntu-rockchip project, would be an apt description (apt, heh).
That's the context I was thinking when I wrote the title, but I definitely see how it could be interpreted in a negative light (not intended!).
@@Level2Jeff I see where you're coming from now. I hang around in spaces where people don't use that word so thanks for the heads up
Like 0Lone I associate AWOL with dereliction of duty so it was confusing for a second between the title and content of your video
Yeah, AWOL is not a good term here.. the video itself is interesting and I tend to agree with the points made, but the title implies a very different stance than is presented.
I'm an embedded Linux engineer and I work with Rockchip SoCs.
Rockchip software support is shit. The kernel they provide is years behind the mainline kernel and it's full of bugs. Not only do they make no effort to upstream support for their chips, they actively make it harder for those changes to get up streamed.
Apparently there is some premium software offering, but if you email them about it they don't reply. How that complies with GPL I don't know.
Basically avoid unless you're being paid to do it. If it's not worth Rockchips time, it's not worth yours
Well said. I did get an orangepi a long while back but poor support left me frustrated (I didn't know about Riek's work). I can tell you I won't be purchasing any more Rockchip hardware as long as the industry doesn't bother to finish their own software end of the product. Ridiculous they'd expect someone to do it for them for free.
thank you for sharing
stuff llike this will eventually affect arm in general
Try being beholden to nVidia's SBCs and associated products. I'm not sure they'll EVER move on from kernel 4.9.
They comply with the GPL by not existing in the US.
I was about to get an Orange Pi 5 Max, I guess not anymore...
Side note: This should be on the main channel for more to see. It is an incredibly important topic.
This!
Ultimately, if the hardware vendor doesn't want to deliver Open Source drivers to the Linux kernel, you're in for a major kicking. At best, you'll need to reverse engineer with no assistance... at worst you'll get legal threats.
The number 1 advantage of Raspberry Pi is that they pay an Open Source vendor to produce Open Source drivers. It stuns me people buy other boards and expect some random person to spend their life supporting your stuff.
Random people supporting your stuff is the whole value proposition for most users.
Yes, idealistically it’s the ability to modify things for your own needs but the more popular the product, the less people there are who do that.
Collabora has upstreamed an OpenGL driver for the RK3588, and I've been told that a Vulkan driver is coming in a couple of months. There's a big chance that you can run the RK3588 with a mainline kernel in 2025.
There has been one very toxic software company who charges most ARM SBC manufacturers to deliver software support in the form of a working OS with regular updates for a limited time(like a contract) and doesn't allow competition to grow up freely, they always try to destroy rep of anyone trying to copycat them or doing it for free like most people do with these non rpi boards.
It's not an easy job to do as a pro bono work for the open source community and there's always some annoying users who purchase a certain SBC board, constantly demanding new features/updates like you were Microsoft.
Not necessarily always about the want. There are also stupid IP laws.
My personal opinion is that the board manufacturers should take a more responsibility for the development of drivers for the boards they sell. Many times they advertise features for the boards that are never fully implemented because there isn't enough support from them or the chip makers. As a customer it can be very difficult to sort through which features a board ships with versus which features are technically possible. Many people are afraid to touch anything that is not a Raspberry Pi because they have been burned by this. If Rockchip makes it difficult then don't use their chips and stop passing the pain along to the end consumer.
The quickest way to burnout (that I have learned the hard way), is to turn your passion into a job.
👀
i agree
Yeah, but if I hadn't done that I would be a retired factory worker or tractor mechanic instead of someone who sold a successful software company. I'm not sorry.
That can be the exact same thing if you reverse it. In fact much more so. I'm honestly surprised if you don't know anyone that has ie. Inherited business they don't care at all and have driven business down. This is very common and stress is common factor on all of these. On the opposite people that have good careers are virtually ones that discuss about their profession in their free time. I pretty much dare to say bluntry that you absolutely can't exel if you don't have interest in subject.
worse if you aren't even paid for it, and get entitled foss people being jerks
"We don't have to do the right thing by you", "Ok, I don't have to anything about this project anymore unless I feel like it. Right now, I don't feel like it.". No one should be mad. He's playing by the rules. It remains to be seen what happens down the road in a few years.
As a contributor to multiple OSS projects, good on him. Far too often big business abuses the kindness of the community. Quite frankly, the biggest issue with rockchip going back like 8 years has been software support and a reason why I never recommend their products.
Back when I first got into Linux, there was still a free software/ open source split.
I think we need a bit more free software politics to push back against this sort of thing.
That needs to be a T-shirt (if it isn't already).
Don't like my code? Fork it!
I want one!
Not sure if the art of it should be a split arrow ↔️ or a fork 🔱 stabbing some lines of code. lol
Some people seem to think that forking is bad, but really it's often healthy.
I rather would like to see, Fork it and contribute back.
I'll definitely get one of those if it is available.
Side note: there is no free meal in the world.
"The best FreeLibreOpenSourceDistro goes AWOL"
Donations in the bank account: $0
I genuinely hate how people demand things of open source developers. I’ve seen it happen too many times and it’s downright terrifying how many amazing developers get stretched to there limits just because people are abusive.
As someone that has contributed to opensource projects (in a minor way) and put out a bunch of "source available" WTFPL licensed code ... if you say that "X solves Y problem" and it doesn't work people are going to be upset and tbh a lot of the tech community in general doesn't excel in the social skills department. I have experienced this first hand, spent a bunch of time solving the problem (there wasn't one really) and then never got even a reply or thanks for making it easier for them.
Share same feelings as yours and been there already and quitted working with them 2 years ago, precisely for those things you're mentioning here.
And this is reason why most open source projects have no documentation. Your project might be the best thing since sliced bread but it's useless if only you know how to use it.
"have we learned _nothing_ from the XZ situation" comes to mind
Folks are quick to call out “open source theft” but not quick to support the originator. We see it a lot in the 3d printing world. Companies profiting off the hard work of devs who barely make ends meet. Thanks for making this video. It’s sad to see this happen but I know all too well how they are feeling and don’t blame Josh at all.
I did a open source project for a certain 3d printable storage solution. The amount of comments on Reddit that said to support the developer and how much awards those gathered was a bit heartbreaking when nobody actually donated. Sure, it is a bit rough around the edges, but there are a dozen daily users...
@ folks are great about talking a good game in the 3 d printing world. They are quick to attack folks for perceived license violations, but actually supporting folks in ways that matter? Crickets. It’s disheartening some times. Sorry you have fallen victim to it.
Putting corporations at the same level than individuals in open source licensing is the Trojan horse that will eventually destroy a project. Change my mind!
That's a legal landmine. If you're building something of note it's better to relicense under GPL to prevent this stuff in general even from individuals
its more deeper than that, open source licenses work almost like a peer to peer model in which eventually you get leechers that only take and don't contribute. Free software licenses are even worse, they put the user needs before developer needs, leaving everyone unpaid and scrapping by.
Yes, I know those licenses require code to be contributed back, but code goes in, it cost time, resources and money, nothing comes out.
Open source ends up being donation of time and money, hopefully you get code contributed back, but its basically a give away.
That's a terrible business model, doing charity development of software.
So, no, even if you changed the license so only individuals could use the software, there's still this problem that the developer never get anything back for their work, except the satisfaction of making a good software (because that's almost impossible in commercial settings)
Maybe we need a new license model for open source software that's not a "free license" of use, something that require users to pay back in some crypto-currency for the use if its not a personal use.
Nothing is free on this world, so why open source software should be free to use ?
I want a free-license power generator, and chip foundry, and sand mining, if everything was free-license and resources were free to take from the ground, then maybe it might work.
Define corporation. Define Individual. In any case, most Open Source is built by a Corporation (RedHat, Intel, or Google)
After I said that, I read the Drupal , that's neat way of solving the problem, track the leechers and make it very public that they don't contribute. Maybe shaming works, and incentivizing the good contributors is a good idea, hopefully the pool of resources donated to the project goes to those who are maker and not to the takers.
I still would prefer that sort of thing in the license (somehow), not as policy of a ONG that controls the project.
Monad just re-invented the concept of selling software. But now “with crypto”. Wow, what a novel idea.
Knowing how engaged Joshua was in the community it just saddened me to see him feel so bad after all the good he’s done. I mean I have a Rock5B running rn with his image. He really is someone like no other taking up a task as big as he did and I hope he can find something that’ll give him the same spark initially
Literally one in my shopping cart right now. You saved me $150
Technically the dev did not go AWOL because they announced they are leaving.. It would have been going AWOL if there was no notices of it or anything of the sort
AWOL can just mean 'deserted' or 'absent', though it often has the connotation of being unannounced. I generally hear it as "being gone abruptly" though, so that's the sense I'm using for this title.
Technically the term originates from not being given permission to leave. If someone goes AWOL, it might have been unannounced or they could have said they intended to leave. The problem is that they do end up being absent, and in terms of military service, or a job, it's not going to be a good time when they get back.
@@eirinym Indeed the question here is who must the dev seek permission so they have leave? I'd say AWOL could apply but really shouldn't as the dev you can argue works for the community, so they don't get to announce they are taking a break without community permission perhaps. But really I'd say the Dev has the ability to give themselves as much leave as they want - its not like they are working 9-5 for a regular paycheck being an open source developer.
It's absent without official leave. A military term.
In short being absent without permission. Since this developer does not need permission to stop he cannot be awol. But it's fine if you take the more generalized meaning. But even in that case the announcement, makes using the term stretching it.
Smacking my head at this thread. Good grief.
Interesting to hear about your experiences and setups. Right now, I am just barely skimming a burnout, mainly because I am the only developer in my company and have to develop a fullstack app because "the boss said so"... And, without me really knowing, went and advertised that fact. So now I am literally crunching to meet early december as a deadline...kool.
Thanks for your insights! I plan to be more active in OSS and FOSS one day and write some fun software I need myself or could be useful to others - so, hearing how long-time-doers do it, is very insightful. ^^
Oh boy, that's a tale as old as time :O
The crazy thing is, I've had some of those situations turn into a nuke that destroyed a project entirely... and others turn into opportunities that went on to be amazing for the startups-it's probably more often the former, though.
Especially when you're solo, I think a lot of management doesn't realize just how low the 'bus factor' is. If you walked away today, would they still be able to launch? If not... that's a massive business risk for them, and they pay a lot for insurance to mitigate fire, or theft, etc. Why can't they invest in software? :)
If the bussiness depends on you they should pay you accordingly.
If not, that's exploitation and you should say bye bye to those abusers.
Time to ask for a raise.
@@Splarkszter Easy to tell others to do that. Hard to do it when your in the trenches. The job market for devs right now is rough.
It's why they invented "quiet quitting". Do what they pay you to do but no more. I'd be looking for another job at the same time. Work your 8 hours then go home. That's how you avoid burn out. Tell them early and often that its more work than one person can do. Do it in emails so it's documented. Then when it goes wrong you have have a paper trail.
The reward for hard work..is more work. If you go the extra mile, that becomes the new normal.
"If you don't like the way I do it. Fork it!" will be my official response to anything anyone complains to me about.
Indeed! More would do better to stick to this.
@@JollyGiant19 That would be awful.
@@eat.a.dick.googleOnly because you've never been on the receiving end of a spam PR, not to mention ones that just rewrite everything and demand you to merge it in as is.
Rockchip is a good example of the lack of software support. I bought some Raspberry Pi alternatives with the RK chips that are gathering dust at the moment. Nice chips, poor support.
Every time I see a video like this reminds me why I avoid ARM computers and I stick with x86 as much as I can:
- no standardization whatsoever, I have 15+ years old x86 devices, long abandoned by their manufacturers (or having their manufacturers gone completely), that can still boot the latest vanilla Ubuntu or Debian (even Windows 11!!!), good luck doing the same on ARM unless some burned out unpaid volunteer decides to do something in their spare time.
- proprietary boot-loaders and kernels that get abandoned as soon as the device is out the door. Does every ARM computer really needs it’s own proprietary bootloader, top secret encrypted device tree and so on just to boot?
They can’t even be bothered to submit patches upstream to the Linux kernel or even to Armbian. Make these devices be usable longer? Oh the horror!!! We can’t have that!
The only way to even use these things is to use the manufacturer’s bespoke OS, which they never bother to upgrade at all and only works on specific versions of their computers. Good luck even running the manufacturer’s software on different generations of the same hardware.
Can’t they get their stuff together and figure out on a standard way to handle booting, hardware detection and resources allocation? This is a problem that has been solved for over 50 years for crying out loud.
People like to hate Microsoft (with good reason for all the shady things they’ve done), but for all their failings, in the 90s and 2000s they forced some standards on all x86 manufacturers and now everything simply works (most of the time anyway). Maybe the ARM ecosystem really needs someone like that to strong arm them into compliance.
ARM people like to talk about how ACPI, UEFI, BIOS and other x86 related things are bad and ARM is so flexible and superior. Sure, x86 and it’s associated technologies are a big pile of ... antiquated stuff that really need a thorough cleanup, but they are STANDARD for every x86 device, from the lowly industrial board hidden in some old dusty CNC machine to the fastest supercomputers ever built.
I had some hopes the Risc V people, having the chance to start from scratch, would avoid these issues, but after watching the latest development in this field, I’m not holding my breath...☹
RISC-V maybe?
15 year old devices are a real energy hog. You could buy a Pi (other SBC are available with less support) with more CPU power for $55, use a 10th of the electricity, and get more performance as well. There is a law of diminishing returns on keeping older kit running. And of course, the Linux kernel keeps dropping features it thinks no-one uses anymore, so even those old machines will stop working soon. Its worth doing the energy calculation.
What’s interesting is on the “server” side there *are* standards, which is why we can boot the same base distros on Ampere/Graviton/Grace. It’s the desktop and mobile/embedded SoCs that are a mess.
This video was eye-opening to me. I was confused on how to do open source properly. But the drupal example really showed me how to fix the maker-taker problem. And what open source projects I should look out for, because I want projects that protect themselves from the burnout problem you describe
At the very least I will start donating money to the FOSS I use day-to-day once I get a job next year
And also see if any of the projects you use have in-person meetups or conferences. I try to attend some every year, it helps to keep that community connection alive!
Lots of respect to you for taking up issues with burnout and the beautiful but fragile structures hat keep Open Source alive. All the best, Per (DK)
I get it, it’s a lot of work maintaining something of this scope. I’m a Debian based distro guy but this is still a huge loss.
Hello Jeff. During my career I helped with many Open Source projects and at one time was a contributor to Drupal. What I call my greatest accomplishment was forking Cygwin to create MSYS which someone else forked to create MSYS2. It's easy to burnout if all you do is Open Source so adding other interests helps. If you feel burnout, back away for a few days then come back to it.
As a biologist I must say that parasites are part of the ecosystem and we must accept that... that said, it is also part of the ecosystem to build the best possible protections against them.
I've tried to shift my thinking away from burnout or success/failure and more just letting my interest in things run its course. It's not that I failed at some project I undertook, I took the project to the extent I was interested in, at least for now. Allowing myself to emotionally distance from projects and then come back later is very freeing. This doesn't help at work much, but my advice is to talk to your manger if you're experiencing burnout. They might be able to help!
Companies need to be prepared to fork the project in accordance with licenses or write their own code if they make the devs mad. If enough developers exert their control over the software (by not continuing maintenance) then perhaps the companies will analyze the risk of losing the developers and find ways to support the project.
Absolutely. A maintainer’s quality of life should always trump outside demands.
Sadly too few adhere to that to try and do right by their users and while noble, it’s only worth it to do if you directly profit in cash. Exposure kills!
An open source developer is a bringer of gift. If you are one, remember that, don't be scared of firing your users.
I've dealt with burnout in the past at a prior job and to a lesser extent at my current one. What made the real difference was being in a role that I wanted to be in, with potential for both personal and professional growth. To prevent it, being surrounded by passionate, caring coworkers who prioritized mental health, pushed for reasonable deadlines and workload, and actively encouraged us to say "no" to additional projects and duties without discussing amongst the team first.
I recovered, and due to my team's culture, we prevent it from taking hold in anyone on our team. The community we surround ourselves with absolutely makes or breaks burnout, and to see such a great contributor experience it due to unreasonable demands is such a loss.
Thank you for bringing attention to this issue, Jeff. It's truly crucial.
Jeff, Truly a sage take on the whole industry at this moment. The sales, marketing, manufacturing, megastructure of capitalism fails to feed its golden goose - it's a dynamic I have seen in the industry for as long as I have been participating (Since `96). How to pay for development? How to balance the demands of business and production against the creative industry of coding. In hospitals they routinely do not pay healthcare providers for time spent writing patient notes and communicating with other providers - too detailed a note takes too much time to write and read from the perspective of the business. Fundamentally writing code takes thought and time, and business still doesn't want to take "longer than necessary" to do that task. We all have heard the "Ship It" mantra. Where what makes for good code and a good tool is thought, refinement, followed by more thought, and more refinement.
Business wants "something to sell" and at some point doesn't really care what it is. That's not healthy. If the business apparatus can't see fit to work with its builders and nurturers, then strong boundaries, and stepping away is the only solution.
Supporting software is very difficult, one thing I learned from several decades of doing it.
BTW I'd love to see more about running Linux other than Debian on ARM , especially Rocky Linux. Much better than the Debian side IMHO
just do it yourself with rocky linux
When you use open source you absolutely are not entitled to any support. The audacity of people making feature requests, there shouldn't even be request trackers for open source software. You want something fixed, you do a pull request.
@@superchiaki 100%. I was just thinking that, there was a fedora that could run on a Pi, so there is a template to go by. I am now a retired software engineer and I spent a lot of the last couple decades building custom Operating System distributions and porting systems from one OS/architecture to another. I've upstreamed open source before to several projects, I have considered throwing in my hat do do more.
@@monad_tcp Well, if you do a pull you should push it back up. Just whining is not helpful, but many don't have the skills to work on software. They should not be cut off as long as they are giving useful feedback on problems and feature requests. Maybe if you can't support the code you should kick in a little bit of money.
This video proves that you've got good ethos in your work but also that you're a excellent person all around. Good on you for speaking out about this. Open source projects need support, collaboration and kindness, not nastiness and gatekeeping. I agree with Veronica (Explains): Linux is awesome, and so are you.
To avoid burnout is to avoid falling into voluntary slavery. In tech that means scheduling a fixed 1x24hrs day off every week and having something else todo that has large non-tech content such as charity.
It's getting harder and harder with security patches now considered as important issue. In the past, volunteer developers can work at their own pace. Nowadays, it's becoming dangerous not keeping the software updated.
@@bltzcstrnx If it's that important, then they'll provide support for it. There's a lot of times where people will scream about urgency of something but if they're asked to help, suddenly the urgency fades away. Unless you're paid to be on-call 24/7, don't be.
I recall there's a cdn provider (that's open source) that provide package distribution service that people use on their website, like jQuery, well, the cdn went down, the moment the creator went to sleep, and the cdn breaks tons of website, it went down for 13 hours iirc, so basically that xkcd comic lol
With regards to avoiding burnout in my career as a computer engineer before I switched fields due to losing the love for doing it 9-5, was by scheduling my time in a calendar and sticking to the time I alotted to various things I needed to get done, and wanted to get done. This allowed me to stay on track and make sure everything got the time it deserved and when it was time to walk away at night and wind down and spend time with the family it was considered sacred and was scheduled. No one is useful when they aren't at their best and burning yourself out only works against you. We aren't machines so we can't try to keep up with working like them. Finally, in my case I ended up Paraplegic at 42 due to an unfortunate accident, and the first thing I thought was, thank goodness I took time to spend on doing things I wanted to do that I can't do any longer because of my impairments and pain levels. My life changed forever in the blink of an eye at 42, and we need to remember something like that could happen to anyone and it really puts things into perspective. At least hindsight was 20/20 for me after it happened to me. Cheers!
Love the Cardinal Glennon shirt Jeff! You forgot to mention your health issues you have faced over the pasted few years that have also added to your plate when building your base .
I have always seen the authors with utmost respect. And always have been humble and thankful when asking for something.
Unrelatedly - Cardinal Glennon represent! I spent entirely too much time there and remain thankful for all of their expertise.
Thanks for the update, I think this repo is invaluable and I will have a look into how I can support it better.
Our circles overlapped years ago when I managed drupal for Canonical. I shifted my focus and became an open source lawyer. If you ever want to do a collab and talk about open source and legal stuff, I’d love connect. We share a similar philosophy on this. And I too have seen and felt this burnout issue.
In February of this year I was quickly, ludicrously quickly, heading for burnout at my IT job. Most of the burnout was being caused by a certain person from a team I left due to toxic, toxic behavior from them. I'd finally had enough and resigned my position and took many months off from work. Only way I could survive.
If you care about money, don't do open source. It's simple as that, when you start making open source projects you are aware someone might take it and make millions off of it and give you nothing. If you can't deal with that, don't make the project open source.
Thanks for the shoutout, Jeff!
Thanks for building Drupal! :D
I used Citra a bit, but mostly gave in support time and effort
And then the lawsuit happened, but I am happy I helped propel that project on, even if for a bit
this stuff is never talked about enough so thanks for doing that
Don't forget about Terraform and Vault changing licenses, too. Those have indirect impacts but affect a lot of people.
Just wanted to thank you for your videos and all your hard work :))
I've made a few open-source projects that have become popular, and I agree wholeheartedly with your video, especially around community. I think as well things that are used by the end-user are a lot more challenging, because your community becomes largely comprised of people who aren't necessarily technical. It almost becomes a chicken-and-egg problem where you want to encourage contribution by making your app better, but at the same time you don't have the resources (especially around UI, testing and localisation) to accomplish that. OSS contribution doesn't have to just be about contributing code - I am very thankful to the graphic designers, translators, testers, and community managers who donate their time to ensure developers can keep developing.
My full respect and support to your words! And, hey, thanks for you're doing ✊
It's great hearing your view on its all. Burn out is real.
I like to establish a philosophy around each new project I start; and commit myself to it it. Otherwise if something gets popular, it's easy to begin to resent the lack of financial gain or contribution. If you're doing it because you like building things people use, then remind yourself why that's why you're doing it, and be prepared to take a break from the project if it gets too much. Jumping between projects and personal hobbies, even if everything as a whole slows down, has helped me to continue enjoy working on it all.
People "stealing" code or trying to compete with your project, using your own work, can be frustrating to some, but I've found that persistence and maintaining the open-source mentally will protect you. Sometimes you can find more strategic ways to protect yourself here, but most times these stresses come and go without real detrimental outcomes.
Surviving financially as an open-source contributor can be hard also; balancing your personal needs isn't easy, and I can understand why some people might start to resent the situation if they feel they've sacrificed so much without improving their own personal outcome. I feel very fortunate to have the privilege to work on open-source, and not need to stress about finding ways for it to make me money, as I can get by with side hustles and being content living a simple life. I've long given up on some dreams, and accepting things as is will lets me enjoy what I already have more.
GitHub sponsorships and other gifts pay for about 10% of my bills, which helps, but it's still not sufficient alone. It's definitely not for everyone as a result; not without some UBI system at least, and that has its own implications I'm sure.
Very tough subject. My personal approach is to document as best as possible what I’m doing. This way I get way less issues, and new contributors feel empowered taking it over. So I don’t feel too bad stopping working on one of my OSS projects or even giving away the bdfn ("now") title. And as you perfectly explain, ignoring is sometimes the best way to keep your mental health.
Thanks for talking about this. It has been a good 20 years since I've "taken point" on a large FOSS project, and damn do I completely lack any ability to miss it.
That said, there are things one can do to help themselves that should get top billing:
• When somebody is willing to contribute, even if it doesn't fit exactly into the ideal process, strongly consider taking them up on it. EG Many people who can contribute a bugfix patch don't know the first thing about writing tests. Sometimes a clear and complete user story is golden-reserving the community for only those who can write code and tests and documentation is asinine.
• Most people don't want to you to fail, they just suck at communicating it. Keep this in mind and it is easier (for many) to not take things personally.
• Many of the best contributions come from people who don't author anything directly, they engage with others to help them make better stuff instead because that's what they're good at.
The flip side can also be a challenge: it's often hard to donate money to FOSS developers, setting that up is just one more challenge. My current job is shit, be in the past I've had an annual budget to donate to FOSS we rely on. These days it's just my money.
My Ask: please, set up an account somewhere, somehow, that people can send money to. IBAN, PayID, even payal if you have to. It might only be $10 a year, so don't put lots of work in, but put it there just in case.
+1 for including links to materials !
The meat's always down in the description! :)
I think the main cause of open-source rugpulls, burnout, etc. is that there's not really any straight-forward ways to license source-available software that isn't open-source too. If people expect more back from the work they do than the license requires (and then don't get it), that's going to lead to burnout. Or people only can afford to hire a lawyer at a certain size to then write a license that actually expresses their intentions. Having open-source as the only way to license source-available software isn't good for anyone involved, including and especially open-source advocates.
Yeah we have hundreds of SBC now, and so much of them are utter trash when it comes to OS support. I wish companies would quit dropping 10 models a year with piss poor support (Orange, Banana and Friendly are notorious for this). Its also something thats often missed during reviews of new product because the reviews are so shallow. They create a baseline build of an OS and then just abandon it. I am not a PI Foundation fan because of the way they have acted over the last decade but at least they have enough support to have a continuous release of updated OS's. I'm sorry to hear this dev got burned out. With 31 years as an IT professional, its a sad norm that most of us will experience.
Interested to understand the phrase " because of the way they have acted over the last decade". The only thing that comes to mind is keeping business alive over covid/silicon supply crisis by prioritising industrial customers (and by that I mean any industrial customers, not just those buying thousands of devices)
@@jamesh9756 When I first started using Pi foundation products their focus was ostensibly education and hobbyist. They gained a lot from the community that sprung up around their products, in part contributions, code support, projects built to take advantage of their platform that also drove a lot of interest, etc. all while maintaining their low manufacturing numbers that led to constant shortages on launches and for a lot of each product life cycles. Some of us had great ideas, and wanted a way to purchase bulk and implement and sell completed devices and projects on a small business scale. When we tried approaching about increasing purchase limits even in novel ways (like having to pay for matching units for charity) they said no. They said their focus was on the individual. So we lived with that. They still grew in popularity in large part due to the community. Then they made the call to focus on industrial and commercial customers. As someone who has contributed to OSS at times for a company to be all, nooo really we are about the individual... so individuals contribute while they grow, and their product and reputation gets enhanced..... and then give people the finger and say, nahh just kidding mate, we all about the industrial customers, but thanks for what you did to get us here. Its as much their handling of their customer base, as the proliferation of cheap MCU that is why there are so many competitors now. So yeah its about the attitude perceived about a business. Like it or not, I see this as them saying Let them Eat Cake. So I avoid their products now. I will gladly burn 100 hours doing compiles and making a competitors board suit my needs.
My opinion. Not looking to argue, just how I feel after my dealings with them over the years.
This might be my favourite video of yours. It's very candid, and honest.
I also ran across you a few times when I was also involved with Drupal, and agree with your thoughts.
Burn out is real, the maintainer for ubuntu rockchip was very kind, so was really sucky seeing him suffer.
“If you’re unhappy, fork it!” Is a fantastic motto.
brb, I'm gonna go sell your ansible roles /s
(btw Thanks for those. That's how I found you, actually. Though I thought you were older when I originally found your ansible roles lol.)
edit: Also I still don't really know what Drupal's good for personally. It's never made a list of ideas or platforms for projects (partially, likely because I don't like to do PHP that much.)
Drupal used to be great for blogs, small mom-n-pop websites, even small catalog sites and medium business. They moved upmarket 5-10 years ago and now it's mostly for custom enterprise CMSes. A lot of big sites still run on Drupal, and my blog does too... but my blog is a horrible use case for it (IMO).
It's similar to Wordpress, lets you quickly spin up rich dynamic websites, and has premade apps for common use cases. And it can be customized in PHP when you need to, without needing to write the whole website yourself.
All of my experience with it is customizing and hosting a simple Drupal website 8 years ago.
That post (along with everything I know about ARM SBC’s) is why I will only buy RPis for now. You can make banging hardware, but it’s a paperweight without support. It’s why Apple laptops could get away with 8 GB of RAM for so long (literally until they needed to run AI locally)
I never knew about this channel until this video... 🎉
Joshua is a saint, he did not leave us, he ascended
Changing license is fine as long as it's done legally. The old versions that were released under the more permissive license can still be used and changed under that license, only the newer changes are more restricted. MIT -> GPL or MIT -> proprietary/source available/restricted can be done without too much effort. The other way around requires a new license from all authors / copyright holders. But either way, it's usually causing community outrage, because people don't like when they used to get something for free, and they no longer get the thing for free. (Myself included, moved over to OpenSearch and Rocky Linux.)
My software philosophy is similar for my own code that I write as a hobby: If I'm giving it away, I'm doing so with no strings attached. If you can make more money from it than I could, good for you. I don't care about attribution either. So it's usually public domain, Unlicense or similar.
Changing from FLOSS to proprietary is bait and switch. Why don't just make it source-available or proprietary since the beginning. Unreal Engine do that, and they have many contributors. Personally, I don't have problem with developer choosing their own license. What I don't like is the recent bait and switch. They use open-source for clout in the beginning, then change the license when it got big.
Ngl sad this isn't a main channel video. I think the topic is pretty important.
Do you usually lie?
@mavfan1 some jokes can be seen as straight-faced lies
The distinction on my side is mostly on production effort and less topic... I figure the YT algorithm decides what it wants regardless of where I post the video these days :(
Rockchips biggest problem is its inability or willingness to support its own chips. On paper the chips are great. But they simply do not provide even reference level drivers for the hardware they make. How do they expect people to use there chips on mass scale? Now that the key open source devs are walking away because their work is ether stolen or not appreciated. .. Rockchips going to have a sobering moment I think.
Also the whole makertaker situation is one of the reasons I'm reluctant to release my code. Always wary some one taking something I've put my heart and soul into being taken by some one else and making money off it.
You should hear Brodie's (with his Tech Over Tea podcast) interview with Michael from Futo. He sounds about as ideological as RMS, but focused on making sure Free Software developers have a reliable source of income, and that seems to be the raison d'être of Futo, which includes exploring alternative licensing that can still give users all the same FSF freedoms, but can help guarantee funding.
Pretty incredible what kind of a thing some people expect
This is a old war, because of licenses the Unix have been "forked" and now we got Linux... the future will be divided into full open-source/hardware or full locked systems 😮
In case folks don't know, "🚩 The Anti-Capitalist Software License" is a thing! Not sure it would help in this particular scenario, but... it's a thing to be aware of for those putting new projects into the world, and/or considering changing licensing for their existing projects.
more people should be looking at the projects/people who DON'T have messy drama and asking what they've done to avoid it
I don't. In the very early 2000s, I wrote a bunch of perl libraries for music tagging that were useful in my personal project. It was a huge mistake. Yes, lots of people used them, to my surprise, but I didn't have the temperament, the time, or the social skills to maintain these libraries.
Creating an open source project is, I discovered, a way to turn your hobby into a job, and when you have pathological demand avoidance, a job sucks.
Yes, people burnt out from greed and power, most of the time bad people are stronger.
Why is support on arm so spotty when anything x86 just seem to work from the get go. Does Intel and AMD do the work or are arm chips so different each one needs work.
X86 is much more standardized with ACPI and all that. Very little effort outside of drivers needs to be put in for it all to just 'work'. Arm devices are all slightly different and custom so a lot of the time you need to do a lot more than just drivers to get it to even boot Linux.
when arm v6 came, everything had to be rebuild, same for v7, and nothing is that much diffeent now
every company does their own arm stuff, every couple generations everything breaks and has to be rebuild for the new chip
x86 is big and heavy because they dont do that, they keep the old modules, the old stuff and add on top, this is why arm can be ligther and power efficient, they get rid of the old, and invent again, but forget to offer drivers
on x86 alot of 16 bit stuff still works, up to 64 bit, all is there if the os can handle it
x86-based systems are almost always PCs and PCs have been historically standarized when it comes to their pre-boot environments: UEFI (BIOS in the past, though many UEFI implementations still offer 'legacy BIOS', so it is possible to boot old, good MS-DOS on modern PCs), ACPI etc. That is, standard interfaces exist to provide OS loader/booting OS kernel with basic HID, I/O, video and means to obtain info on available hardware. Therefore, on x86 generic OS images can be used, because no prior knowledge of hardware is needed to blast off. x86/PC is rather unique with this luxury of freedom of choice. ARM and RISC-V worlds are much more locked-down and in the segment of SBCs practically devoid of standarized preboot environments.
This is good advice for all professions!
I'm not good about it. Lots of folks need help, but it's not our responsibility to help everyone. That's just reality.
I could, at the expense of my family, health, and sanity, do pro bono work 40 hours a week (on top of income-generating work) but that's just not realistic or healthy.
good info thanks for speaking on this
This does not go nearly far enough. Desktop Linux will never become a thing without consolidation, and thus also bundling of forces. Because confusing and overwhelming the customer is poison for the success of the product. - As the saying goes: “If You Confuse, You'll Lose”
I’m big into the DIY eurorack community and the owner/proprietor/sole-employee of Mutable Instruments, the biggest inspiration to the entire eurorack community, completely stepped away from the game. She made her designs completely open and a year or so before she closed up shop she was invited by Arturia to corporate gathering and 3 months later Arturia released a product with her code and used her brand in marketing with zero compensation. Her products were then given the same treatment by Behringer without even getting the free lunch. I get that the license doesn’t require compensation but then there’s also just being an asshole. Then we don’t even get whatever these leeches do with our improve things because they won’t open source their product.
I think open source is the reason we have nice things but when that ecosystem can’t actually be fostered I genuinely wonder if we’ll continue to have nice things.
Korg has been caught using the Linux kernel without providing their changes, as their OASYS, Kronos and Nautilus keyboards are Intel Atom PCs (which can be upgraded like PCs, since it's a bog standard Industrial ASRock motherboard inside the case. Yamaha does use MontaVista Linux in some of their hardware and you can request the source code if you want.
I’m not in tech. But burnout is real in everywhere.
Great insight, I feel like I learned from it.
They forgot it was supposed to be "Research and Development" not "Rip-off and Duplicate" ... So frustrating!
But if you give them explicit permission to do so, why is it wrong?
No one is hurt, except for egos.
Now that our house construction is finally finished, I finally got around to trying out the Orange Pi 5 Plus 32GB that I bought months ago. And what did I have to realize? The hardware is great, but the official software support is total crap. I am extremely grateful to Joshua for his work. Without his valuable work and time invested, the hardware is almost useless. I completely understand his reaction. It's a bitter loss and I really hope he comes back. But your own health comes first. You do something in your free time with extreme dedication because you have fun doing it. At some point, those around you see this devotion as normal and begin to make demands. The tragic thing about it all is that this is not limited to IT or open source topics, but happens everywhere. I experience exactly the same thing in my IT job, but also in my private life. At some point it just doesn't work anymore and you need a break or stop doing what you actually love completely because otherwise you'll just break down.
Get well soon Joshua!
Yes I agree, open source software has given us a lot of nice things and opened ways for people to widely benefit off of each others work. But the sense of entitlement that we as open source users can have can be very bad.
"...and if you don't like the way that I do it, fork it!" I like what you did there. 😂
Also, I really like the liberal with your no's bit. I'll have to try that. Thanks!
This should be on your main channel
I wrote a telescope driver for the Meade brand of scopes. I did it as I had one myself and wasn’t happy with what was available. 5 years later, I’ve found myself in the position that I don’t want to use that scope anymore, have an different one that works better for what I like to do. But, I’m now stuck supporting this code that I wrote. It feels more and more of a burden, but I don’t want to abandon the project until I don’t have access to the scope anymore.
I think you should take @dareelcatskull 's advice above (in a comment) - don't think of it as failing, but letting your interest in that part of the project run its course. Make all the artifacts you use to create that driver (no matter what state they are in - no don't clean them up!) available, and step back. The only one trapping you into supporting that code is you.
The more software ecosystem issues I encounter, the more important standards seems to be, having communities build with the same compatibility requirements seem to be the only alternative to high resource businesses
Open source foundations are a good idea. Basically they do the work of canvasing for donations and support from companies and then redistribute that money to projects that need it. Specifically to avoid the XKCD situation where a single developer is propping up a huge amount of downstream development.
The only reason his work is awesome is support for hardware acceleration for jellyfin/immich.
Crazy how RK3588 spec sheet of encoding/decoding reflects the actual use (FINALLY) using his stuff.
It makes it an option compared to intel N100 for jellyfin specifically.
What is not fair is when they take open projects add to it and then close it so no one else can make from it,, keeping all the spoils for themselves. Talking about burn out do you ever see your family? You run two channels, do projects, write software and up load to github and build a studio workshop.
One solution to this is dual licensing. As copyright holder, you allow anyone to fork your code and use it without restriction as long as changes are fed back to the community (AGPL rather than GPL, so people can't avoid releasing the source code by keeping it only on their servers and not technically 'distributing; it) and a second commercial license for people who don't want to contribute their changes back.
The problem with dual licensing is that you need to either have a single developer, who re-implements any external pull requests, or have a contributor license agreement in place, which assigns rights to the copyright holder. In either case, it's best to be up front about dual licensing at the start of the project, otherwise people will feel cheated if yo try to introduce it later on.
Compartmentalization helps me there - it was a lot harder before I had a separate studio to do the RUclips stuff! It's always tough to keep away from the allure of 'just working on that one little issue in that one little project' (which always turns into like 3 hours debugging lol), though.
i think missed a major factor in Joshua's post... money. he clearly states that hosting the project is expensive and he can't afford it. community and not getting burned out are important, but i am seeing a lot of FOSS people with major projects leave the community as they simply can't make ends meet while working 90 hours a week...
30+ years in professional software development here. I got burnt out on Project work, the deadlines, the testing, the go lives and attendant stress, warranty fixes then rolling into another. My boss thankfully noticed before I had the balls to admit it. I was moved into a senior support role which I love and have been doing for 4 years now. Lesson, I guess, is to speak up if you're struggling and let someone else help out. I was considering leaving IT completely, glad I had a good leader.
You are fortunate to have a boss who cares. I think that is the exception, and not the rule, especially when the head of the company is ruthless... it filters down.. sometimes slowly, but eventually people who last adopt the mindset/approach.
@@johnlaurencepoole6408 Very true. My organisation is ex public service so the attitude is probably more relaxed than somewhere that is private.
I have managed several open source projects and I just handed them all off one day, or archived them because I just couldn't do it anymore. People demanding things, being super pushy and just not willing to put in their own time to add commits but just whiny little asses. Yeah no thanks.
I know exactly how this dude feels, some of us do things for the wider community for free, but never receive the respect or thanks we sometimes deserve, some companies leverage their business models on things found in the opensource community but never give anything back, whenever I see someone fencing some new SOC or SBC on RUclips, I always ask them if the manufacturer has posted the source code to their operating system, have they provided the source code for the driver stacks, the answer is either silence, or its no, we need to stop helping these ungrateful companies and only work on hardware that does provide everything in the open and also assists and gives back to the community.
I get what you mean when you say "you're happy if people can do something awesome with your code" but your followers having to "chime in" (pay you/ donate to you) while these companies making money using your code aren't reimbursing you, means your supporters are essentially subsidizing corporate revenue.
If these companies would simply pay you (even a small fee) for your work, your followers could either safe their money and still get awesome content from you or their money could be used to increase project budgets or production quality.
Either way, they are directly exploit your "free work" and therefore by extend, also the people supporting you.
6:27 kudos mate. 🎉
It would be nice to create an award at least to recognize these people.
Honestly I feel like we need more licenses where it's like "this is free to use and modify, but if you're a business and your company makes over $1,000,000 a year you have to pay royalties" or something along those lines. Open source and free software (both versions of the word free) are extremely important for society, but developers also need to be compensated fairly for their time and effort. So those who profit off their work should share some of those profits.
In other news I probably need to start contributing to open source projects myself. I'm a recent computer science graduate and I can't find work, so now's a good time for that (though I'm also still just recovering from the burnout my education caused).
No clue how to even get started though and not sure I'm ready.
I wonder if a kickstarter could be made by the original developer for "revisions"
Want a newer version...pay $$$ to support the developer...once the target is reached, the revision is released
It's time to "buy them a coffee" to help support those intelligent and ingenious enough to release code that makes all our lives easier
I'd rather send money direct. Running a kickstarter is a lot of work and risks the developer getting even mroe burned out. It's also a different skillset and adding that as another barrier seems like a bad idea.
What each of us can do is contribute money if we can. It makes a difference, both emotionally "someone cares, also someone likes what I'm doing enough to send money", and the obvious "I have money" way.
Rockchip wants to live in their comfort zone.
They don't care about new products because those don't move that many chips.
Their only metric is chips sold. No future prospects, no insight.
The same exact thing Microchip did. Their history is Rockchip's future.
Sent from a Fydetab Duo running BredOS. This stuff is dope.
I like to get myself a very comfortable life with the stuff I build.
Why not other people as well, if they help me with my stuff?
Sadly, this really isn't the case for all open source communities. It's really hard to build nowadays. There is so much competition as well. So much VC money nowadays.
We had similiar stress and loss of motivation with LZMA SDK as far as I remember. This was purposeful work to hijack the project and the world depending on this SDK. Do we have similiar simptoms here ?