My partner is on SSDI for disability. If she works, she will lose her SSDI income, but she's allowed to generate income that isn't work/labor.
She is exploring FOSS as a career path because she could accept donations and that wouldn't impact her SSDI. She understands donations would be minimal, but she's hoping it's a way for her to break into the FOSS scene.
As AMD, Intel, Tenstorrent, and other companies develop better hardware, more software developers will be inclined to design for these platforms, and Nvidia's CUDA dominance could ease over time.
This seems a bit optimistic to me. CUDA is currently the de facto method of utilising a GPU's power efficiently. This makes them an easy choice for anyone with serious compute power needs. The other manufacturers are fighting an uphill battle trying to create an alternative that won't be used until it is definitively better.
It's not "optimistic", it's actually happening. Don't forget that GPU compute is a pretty vast field, and not every field/application has a hard-coded dependency on CUDA/nVidia.
For instance, both TensorFlow and PyTorch work fine with ROCm 6.0+ now, and this enables a lot of ML tasks such as running LLMs like Llama2. Stable Diffusion also works fine - I've tested 2.1 a while back and performance has been great on my Arch + 7800 XT setup. There's plenty more such examples where AMD is already a viable option. And don't forget ZLUDA too, which is being continuing to be improved.
I mean, look at this benchmark from Feb, that's not bad at all:
And ZLUDA has had many improvements since then, so this will only get better.
Of course, whether all this makes an actual dent in nVidia compute market share is a completely different story (thanks to enterprise $$$ + existing hw that's already out there), but the point is, at least for many people/projects - ROCm is already a viable alternative to CUDA for many scenarios. And this will only improve with time. Just within the last 6 months for instance there have been VAST improvements in both ROCm (like the 6.0 release) and compatibility with major projects (like PyTorch). 6.1 was released only a few weeks ago with improved SD performance, a new video decode component (rocDecode), much faster matrix calculations with the new EigenSolver etc. It's a very exiting space to be in to be honest.
So you'd have to be blind to not notice these rapid changes that's really happening. And yes, right now it's still very, very early days for AMD and they've got a lot of catching up to do, and there's a lot of scope for improvement too. But it's happening for sure, AMD + the community isn't sitting idle.
Unfortunately the article of the post directly contradicts your point about ZLUDA improving:
ZLUDA appears to be floundering now, with both AMD and Intel having passed on the opportunity to develop it further
Following the links and searching around, I found this: Andrzej "vosen" Janik, the lead dev, says in his FAQ:
What's the future of the project?
With neither Intel nor AMD interested, we've run out of GPU companies. I'm open though to any offers of that could move the project forward.
Realistically, it's now abandoned and will only possibly receive updates to run workloads I am personally interested in (DLSS).
In any case, ZLUDA is really just a stop-gap arrangement so I don't see it being an issue either way - with more and more projects supporting AMD cards, it won't be needed at all in the near future.
How easy it is to install and configure Rocm and also how limiting it is? I also heard about ZLUDA, etc. and I very much want to pick AMD as my next GPU, especially considering the fact that I am using Wayland, but I think they are still far behind NVIDIA?
Since you're on Linux, it's just a matter of installing the right packages from your distros package manager. Lots of articles on the Web, just google your app + "ROCm". Main thing you gotta keep in mind is the version dependencies, since ROCm 6.0/6.1 was released recently, some programs may not yet have been updated for it. So if your distro packages the most recent version, your app might not yet support it.
This is why many ML apps also come as a Docker image with specific versions of libraries bundled with them - so that could be an easier option for you, instead of manually hunting around for various package dependencies.
Also, chances are that your app may not even know/care about ROCm, if it just uses a library like PyTorch / TensorFlow etc. So just check it's requirements first.
As for AMD vs nVidia in general, there are a few places mainly where they lagged behind: RTX, compute and super sampling.
For RTX, there has been improvements in performance with the RDNA3 cards, but it does lag behind by a generation. For instance, the latest 7900 XTX's RTX performance is equivalent to the 3080.
Compute is catching up as I mentioned earlier, and in some cases the performance may even match nVidia. This is very application/library specific though, so you'll need to look it up.
Super Sampling is a bit of a weird one. AMD has FSR and it does a good job in general. In some cases, it may even perform better since it uses much simpler calculations, as opposed to nVidia's deep learning technique. And AMD's FSR method can be used with any card in fact, as long as the game supports it. And therein lies the catch, only something like 1/3rd of the games out there support it, and even fewer games support the latest FSR 3. But there are mods out there which can enable FSR (check Nexus Mods) that you might be able to use. In any case, FSR/DLSS isn't a critical thing, unless you're gaming on a 4K+ monitor.
You can check out Tom's Hardware GPU Hierarchy for the exact numbers - scroll down halfway to read about the RTX and FSR situation.
So yes, AMD does lag behind in nVidia but whether this impacts you really depends on your needs and use cases. If you're a Linux user though, getting an AMD is a no-brainer - it just works so much better, as in, no need to deal with proprietary driver headaches, no update woes, excellent Wayland support etc.
Yes, I am running NixOS with Hyprland at the moment as a trial and most things were pretty well. I know that open source NVIDIA drivers are crap especially if you want to run Wayland, but I am more interested into the AI/ML side as I want to play a bit with open weight LLMs, and Pytorch. I used to do some AI with Tensorflow, but I would like to learn more about Pytorch.
I used to have an older AMD card in the past that I borrowed from a friend and tried to install Rocm and it was an absolute disaster. That was around COVID and even though I consider myself fairly familiar with Linux and very comfortable around the command line, I didn't make it work back then.
The majority of the opinions I have also read were just pointing out that CUDA is just plug and play and Rocm is a lot of tinkering. And I think I am simply too old and tired of this constant tinkering and I would prefer something that will simply just work out of the box.
I really hate NVIDIA and don't like the company but still consider them with something like i3, just to have some peace of mind and know that everything works out of the box with their proprietary drivers.
When I saw this posted I thought : "Oh no, not yet another post about Microsoft releasing DOS source code" but the article is written by none other than the founder of FreeDOS. Nice, and an interesting read. Thanks for sharing.
It looks like it covers everything I should need and more. I'll have to see if it will connect to my bank, or if only some banks use the protocol, but it looks promising.
Unwarranted advice is welcome too :)
I haven't had much luck with journaling, as I usually don't like the act of writing things down. If it can be done digitally though, it's worth another look, thanks
How to Bullet Journal – a quick intro by the original creator (biggest problem is that it’s been turned into a craft project by Instagrammers, more weight put on picking out the right washi tape than keeping track of tasks)
I selfhost Vikunja on a small 1GB RAM VM. It has views for list, table, Gantt and Kanban (I assume that's what you mean by not manually reordering?) You can setup reminders for your tasks like any project management tool as well.
You can access its web interface via port 3456, so no syncing or external app is needed, it's all browser based. Of course, you can setup a wireguard VPN to access it anywhere.
Support for QUIC and HTTP/3 protocols is available since 1.25.0. Also, since 1.25.0, the QUIC and HTTP/3 support is available in Linux binary packages.
It's a slow system. MAL takes user submissions because they already have a big database and those submissions help filling the cracks. If you don't have a database to begin with it slows things down, especially if these submissions will go by trust-based. Doing this by randoms is also very risky.
However, I'm not completely oppose to that idea because it can be helpful for some areas. My suggestion is start with a basic dog tag system, where the anime name, alternative names, status (airing, completed), season and start date, studios (also licensors and producers), age rating etc. These information needs to be scraped for the fastest way to form a quick database, they are publicly available (even on Wikipedia) so it should be fine to scrape. You can even go full Wikipedia after got only the names. User submissions could be useful for the introduction / summary parts of the titles at this stage. For only names (and basic tags), you can scrape AniDB from this list. It's just a search query so shouldn't be against their ToS.
You can also check Kitsu for ideas, I like their DB request system. Pretty basic but can be done differently with the power of ActivityPub.
I see. I can use the Jikan API to scrape animes and mangas which will take approximately 1-2 days after I get approval then.
Oh and I forgot to mention, the federation part isn't really how people think it will be I guess.
The only federation that will be done will be the reviews, threads and the comments in them. With every anime/manga/vn, etc. being a new community which will contain those threads and reviews. And because of that, I don't really know if this project is something people would want to self host. I guess I could provide full dumps of the database every month or something but I suppose that would be expensive. Then there are images as well, which will take hundreds of GBs easily even in compressed form.
Didn't know about Jikan API. After a quick look at their docs, I think it should be a steady source for scraping.
For features, can be done a lot with ActivityPub. Of course the most wanted features would be a watchlist / episode tracker (and possibly an importing from the lists people already have, I switched to Kitsu from MAL that way) but just thinking about federating the all anime/manga titles with basically their own communities out of the box sounds great. Good luck with the project!
Yeah, though I am not sure how the federated instance admins would react.
I am planning for every anime, manga, vn, etc. to have their own communities. This means about over 100k communities being made in an instant.
Maybe instead of creating the communities in an instant, creating them when user activity first happens would be more fit. But this would also restrict other platforms' users being able to comment on never heard or new anime entries until someone from the anime tracking platform comments or reviews them.
Both have ups and downs. Assuming these lists will be in the code, do you have an estimation how big would that be? If you think they won't strangle the code, just go with it. Something like storing them in JSON and loading them when needed could be better for optimization though.
You can also do some best of both worlds, like not creating the communities beforehand but make the titles searchable from the database open for all users. That might require a bigger traffic from hosting side though, but it should be OK since these will be spread to all self hosted communities.
I think you can also ask some of your questions to selfhosted communities.
Well, since 100k titles would make a pretty big database, at least storing the metadata like years, seasons, genres etc as hardcoded could make it run faster than going full-fledged JSON, at least I meant that. However this will be an open source project and there will be localizations, so now it doesn't look like a good idea to me somehow.
Sorry for the confusion. Don't take my words literally, I'm not an expert on databases. I use JSON for my job and it's enough for me, so use what database your project needs. But by full-fledged I meant completely leaning on databases which may slow some things down for the server side. However I now checked PostgreSQL and it seems much better for large datasets, so I don't think that would be a concern for you.
For hosting images, you can go alternatives like some Lemmy sites do: Mirror everything automatically to Internet Archive.
I think people would want to self host, because they will get all the anime/manga titles with their communities out of the box and can moderate their own sites while their users can react to any other community via federation.
The faircode model assumes that contributions from random outside people are minor and that the bulk of the work is done by the founder(s). To the founders there is little actual benefit from being an open source project, anyway. I can understand the attraction of the model in that situation.
My ideal OSS project would be receiving a steady stream of contributions from a wide variety of people without an elite sub group that considers themselves to be "the authors", which would be obviously unsuited to the faircode model. Sadly few projects achieve that and are largely the work of one person.
Let's not forget... the reason this type of licensing exists is because large cloud providers were taking a large code base and selling them as services . Often, the main path for the creators to make any money from their code is to offer a paid, managed tier, along with professional services. They would end up competing, and losing, against those cloud providers.
Not saying this kind of license is good or bad, but the reason is often not to stop self-hosting or screw contributors, but to maintain couple of the only pathways FOSS can bring in revenue.
WARNING. Everything other than the last paragraph is kind of rude and opinionated, so skip to the bottom if you only want practical advice and not a philosophical rant.
First of all Free Software don't need paid developers. We scruffy hackers create software because it's fun. I have a strong suspicion that the commercialization of Free Software via the businessfriendly clothing "Open Source" is actually creating a lot of shitty software or at least a lot of good software that'll be obsoleted to keep business going. Capitalization of Free Software doesn't have an incentive to create good finished software, quite the opposite. The best open source software from commercial entities is in my opinion those that were open sourced when a product was no longer profitable as a proprietary business. As examples I love the ID software game engines and Blender. Others seem happy that Sun dumped the source code of Star Office, which then became OpenOffice and LibreOffice, but then again companies like Collabora are trying to turn it into a shitty webification instead of implementing real collaborative features into the software like what AbiWord has.
..and back in the real world where you need to buy food. Open Source consultancy, implementation of custom out-of-tree features, support, courses and training, EOL maintainance or products that leaverage Open Source software is my best answer. See Free Software as a commons we all contribute to, so that we can do things with it and built things from it. You should not expect people to pay for Free Software, but you can sell things that take advantage of Free Software as a resource.
I'll check this out later, but I've checked up on Redox from time to time, and they've been making steady progress. I sincerely hope they succeed, because I would love to see this competing with the likes of Linux, instead of other novelties like RISC OS.
Open Source
Top
This magazine is not receiving updates (last activity 51 day(s) ago).