mindcrash

Hardware: HX370, 128G RAM, Radeon 860M iGPU, Radeon RX7700S dGPU, Xbox Wireless Cntroller, 2T + 8T SSD storage

Software (as of today, still making additions and refinements): Gentoo/OpenRC (I don't like systemd), Kernel 6.12.58 with additional module for the Xbox controller, Pipewire+EasyEffects 8, KDE Plasma 6.5.4/Wayland, Steam

Experience: KDE runs pretty stable, and only has the things I really need (and not the things a vendor thinks I need).

The first game I benchmarked today was Doom (2016), which runs smoothly on 90-120 fps on high settings.

The second game I benchmarked today was Indiana Jones and the Great Circle (2024) running on ~56fps on recommended settings on the 7700.

The one game I tried today and could not get to run properly was Stalker 2: Heart of Chernobyl. I suspect that, given the many positives on ProtonDB, that's mainly either a configuration or Proton issue. I'll do some more research and give it another try in the near future. Right now performance drops to 5 fps immediately after starting a new game, and the CPU running on 600Mhz maximum when starting the game on Proton Experimental.

For now I am quite happy with the results, and the fact that I likely finally am able to eject Windows out of my life.

"Article 1 (2) establishes that one of the main purposes of the United Nations, and thus the Security Council, is to develop friendly international relations based on respect for the “principle of equal rights and self-determination of peoples”. The case studies in this section cover instances where the Security Council has discussed situations with a bearing on the principle of self-determination and the right of peoples to decide their own government, which may relate to the questions of independence, autonomy, referenda, elections, and the legitimacy of governments."

https://main.un.org/securitycouncil/en/content/purposes-and-...

And now I wonder what fucking insane manufactured casus belli he will create to invade and annex Canada.

So many people thought he was grand standing and just talking shit. But now it looks like he wasn't doing that at all, hasn't he?

His voters thought Trump would be different, he would bring the troops home, put the homeland first, and that he would fight the Deep State.

In reality, he's building out Imperium Americanum, he is fighting wars without Congressional approval and proper casus bellis, he's not bringing the troops home and it is clear he represents the fucking Deep State even more than any of his predecessors since JFK. Shame on him for renaming the Kennedy Center the Trump-Kennedy Center. Which is absolutely disgusting given the reality of things!

Prime example: Invading Venezuela to steal their oil, just like his predecessors did with Syria (if you do not believe me, look where the US Army is located in Syria, and the prime locations of their oil fields).

Additionally: Trump's United States now has given Putin's Russia and Xi's China precedent to do whatever they fucking want to whoever or whatever. Because who fucking cares about international law if even the United States government, home of freedom and democracy and the rule of law, currently doesn't even give a fuck?

So now fucking what?

(And yes, as you might have noticed I AM FURIOUS AS HELL.)

And yet they'll have to pry my OpenRC, cronie, and sysklogd from my cold dead hands...

That's incorrect. Querying installed apps has been severely restricted (and thus mostly useless) and also requires a special nuclear-scale permission since Android 11.

I am wondering what exploit HSBC is using because I really don't think they are using official APIs for this.

Stardew Valley, which has been sold to millions of gamers, has been created using the free MonoGame engine. So ConcernedApe is giving back to the open source software which made his commercial success possible, like commercial parties should.

One of the things I am currently experimenting with is building out my own agentic/assisted computing environment which instead of extending into Google/Microsoft/Apple owned cloud based services, extend into services which run on my homelab environment instead.

As a simple example: A local model which can hook into a MCP service making it understand calendars and appointments which hooks into my own locally hosted Radicale CalDAV service, enabling me to quickly make a appointment through text (or possibly even STT later). I'm curious how much I can get something like Thunderbird to disappear.

A somewhat advanced example: Another thing which recently popped up as a idea, I'm quite excited about and I hope will work out is that I can teach a model the concept of a "package repository", a "package manager" and "systems", which (hopefully) means I can install, uninstall, update and track the status of software packages on my Linux systems without using the terminal or shelling into a system myself.

Summarized: I think some things Big Tech wants are pretty neat, but I would like something without heavy involvement of Big Tech (and/or subscription based computing) instead.

After Skool: https://www.youtube.com/afterskool | https://open.spotify.com/show/36mIOrFwTKIDER8KF0aGrx | https://www.afterskool.net/

Academy of Ideas: https://www.youtube.com/academyofideas | https://open.spotify.com/show/2dio7KUNuDHErlMumZtNt6 | https://academyofideas.com/

Every model with ~4B parameters runs perfectly fine on even a Geforce 1070 Mobile GPU with 8Gb of memory.

If you have some patience you can probably go a little crazy and run a model with ~27B parameters on a Radeon 890M with 32Gb of memory as well (means you'll probably have to get about 96Gb of system memory if you want to get some work done too, but oh well).

In theory you could even run a model which fits in 64Gb of video memory on that "little" GPU (with 128Gb of system memory).

No, you can't run something like Grok 2 (which has quantified models starting with 82Gb in size and going up) but why on earth would you ever want to run something like that locally?

https://github.com/mattermost/mattermost/blob/master/LICENSE...

This means that you can totally fork and patch any restrictions out of the publicly available source code, and there's not a bloody thing they can do to stop it.

Also, from a legal standpoint the "subject to the exceptions" part is bogus, because it is AGPL -- no exceptions can be made to that license, else it would not be copyleft anymore.

https://www.pcmag.com/reviews/framework-laptop-16-2025

I quote: "A DIY Desktop Replacement Dream Machine"

It also outperformed the Asus Zenbook S 16 in their benchmarks, which also has a HX 370 SoC.

The only thing they didnt like was the battery.

If you do want a controller/node setup you're better off deploying something lightweight like k0s, k3s or sidero's talos (since the latter also works perfectly fine on clusters based on pi 4 and above)

When you compare apples to apples (lol) - PCs to PCs - the current gen higher end Framework 13 and 16 are beasts. They contain the fastest mobile AMD chips on the market today, the Ryzen AI HX 370 aka "Strix Point". The only thing faster than that is the Ryzen AI Max+ 395 aka "Strix Halo", which would probably have some issues in a laptop with this size if you want to make full use of it because of thermals. Just look at the kind of cooler which sits on the mainboard of Framework Desktop (which does contain the AI Max+ 395) to get an idea what I am talking about.

You can choose between no discrete GPU, AMD RX7700S or NVIDIA 5070M; or just purchase them all and swap them around if you want to. (and who knows what dGPU will be for sale in the (near) future... and all it takes for them is to make that thing fit into the Expansion Shell)

Due to the included AMD mainboard you get USB4 instead of Thunderbolt 4, which means little to no protocol overhead when you hook up a EGPU and can run a docked 5070Ti or 9070XT at ~100fps @ 1440p.

You get up to six (on the 16) expansion bays which enable you to swap external ports to any configuration you like. They even have created 2 external SSDs you can slot in on these ports and blend in with the laptop (which come in sizes of 256G and 1T), and released everything required to allow people in the community to create their own ports as open source.

And I could rave on and on...

But again: Is it a Apple quality laptop? No. Is it a beast of a PC laptop? Yes.

PS: If you think Lenovo is the Apple of the PC world due to the fact that they could purchase some branding and design from IBM back in the day you are in for a very nasty surprise. I have all the hands on experience with a so called "premium" Thinkpad worth a pretty cool 3500 EUR back then to write this.

But realistically, isn't 100 hours of gaming a month like... a lot?

I mean, if you have a friend who studies (or studied) psychology just for fun and kicks jokingly tell them that you play video games just about 4-6 hours every day each week. See what happens. (I know what will very likely happen, but I do not want to spoil anything. Do make very sure to tell them that (hopefully) it was just a joke, though.)

For contextual, super specific, super specialized work (e.g. SCX-LAVD, the DirectX-to-Vulkan and OpenGL-to-Vulkan translation layers in Proton, and most of the graphics driver work required to make games run on the upcoming ARM based Steam Frame) they like to subcontract work to orgs like Igalia but that's about it.

Fairphone: https://shop.fairphone.com/home

Widely available in the EU, in the US (and elsewhere) only through:

Amazon: https://www.amazon.com/dp/B0FTZFFVYC?th=1

Murena: https://murena.com/shop/smartphones/brand-new/murena-fairpho...

(Note: The Murena version also comes with /e/OS instead of vanilla Android)

If you're mainly looking for deGoogled phones:

Google Pixel w/ GrapheneOS (Pixel 10 support now in early beta)

Any Android phone compatible with /e/OS;

Any Android phone compatible with LineageOS;

Previously any Android phone compatible with CalyxOS would also be advisable but not any longer due to the lack of maintenance

Tests in the wild have shown that the best performance can be gained in combination with the currently wildly popular enclosures with the ASM2464PD chip from ASMedia due to its really great compatibility with AMD chipsets. This chip has been integrated in the AG02 enclosure from Aoostar (which supports USB4, but also supports Oculink) and the UT3G/UT4G enclosures from ADT-Link (which only support USB4):

https://egpu.io/best-egpu-buyers-guide/

You also need a really fast videocard with > 40 Gbps throughput to make Oculink worthwhile, as even a AMD Radeon 9070XT performs nearly the same (and pretty great!) on USB4 as Oculink, through the Aoostar dock, on the Framework 16.

They never changed.

With the bonus that you also get a set of great (and per fork different yet handy) features.

These include:

Waterfox (Firefox) - https://www.waterfox.com/

Zen Browser (Firefox) - https://zen-browser.app/

Librewolf (Firefox) - https://librewolf.net/

Helium (Chrome/Chromium) - https://helium.computer/

Ungoogled Chromium (Chrome/Chromium) - https://github.com/ungoogled-software/ungoogled-chromium

Also as one of the major players, Vivaldi already made a stand against AI and forcefully including (agentic) AI in the web browser: https://vivaldi.com/blog/keep-exploring/. It's a Chromium based browser with a lot of nice features and deep customization options: https://vivaldi.com/

After skimming the article I noticed a large chunk of this article (specifically the bits on deattaching/attaching drivers, qemu and vfio) applies more or less to general GPU virtualization under Linux too!

1) Replace any "nvidia" for "amdgpu" for Team Red based setups when needed

2) The PCI ids are all different, so you'll have look them up with lspci yourselves

3) Note that with consumer GPU's you need to deattach and attach a pair of two devices (GPU video and GPU audio); else things might get a bit wonky

And now we have:

- A extra toolbar nobody asked for at the side. And while it contains some extra features now, I'm pretty much sure they added it just to have some prominent space to add a "Open AI Chatbot" button to the UI. And it is irritating as fuck because it remembers its state per window. So if you have one window open with the sidebar open, and you close it on another, then move to the other again and open a new window it thinks "hey, I need to show a sidebar which my user never asked for!". Also I believe it is also opening itselves sometimes when previously closed. I don't like it at all.

- A "Ask an AI Chatbot" option which used to be dynamically added and caused hundreds of clicks on wrong items on the context menu (due to muscle memory), because when it got added the context menu resizes. Which was also a source of a lot of irritation. Luckily it seems they finally managed to fix this after 5 releases or so.

Oh, and at the start of this year they experimented with their own LLM a bit in the form of Orbit, but apparently that project has been shitcanned and memoryholed, and all current efforts seem to be based on interfacing with popular cloud based AIs like ChatGPT, Claude, Copilot, Gemini and Mistral. (likely for some $$$ in return, like the search engine deal with Google)

Everybody now is like "Hey, we can take something like Kubernetes which is open source and is backed by a worldwide community, and you know like OpenStack which is open source and is backed by a worldwide community and we can build our own computing platform and deploy services and online communities and stuff on top of that"

And I was like "Wait, you guys are realizing that NOW?!? I've been an activist and part of a movement urging you all to try and be less dependent on US Big Tech and focus more on decentralization for YEARS"

Like you I am really happy things seem to get rolling now, though :)

They have developed a Android app (available through different and aforementioned distribution channels, ofcourse), but they have one for iOS too!

iOS: https://apps.apple.com/us/app/organic-maps-offline-map/id156...

Droid - Play: https://play.google.com/store/apps/details?id=app.organicmap...

Droid - Obtainium: https://github.com/organicmaps/organicmaps/wiki/Installing-O...

Droid - F-Droid: https://f-droid.org/en/packages/app.organicmaps/

And given all the privacy breaches Meta's engineers did manage to succesfully execute on iOS until now, given the sandbox restrictions, they are in fact pretty creative if you ask me.

Don't believe for a second you are safe from these fuckers on either Android or iOS.

And trust me, they are running a lot of public and private repositories.

And there are many more orgs and govs throughout Europe doing similar things because there's a (growing) zeitgeist here that the Trump administration nor any American SaaS company can be trusted. This started, by the way, after Microsoft suspended the ICJ from using Microsoft 365 on orders from the White House.

Likewise the Windows 11 (which is just a rebranded Windows 10, just look at the full build number which should start with 10.x) kernel could boot systems from ~2017 onwards. Maybe with some kernel features disabled which most (if not all) Windows 10 users would not miss anyway, but it could still boot without any issues. Those running a Rufus-patched Windows 11 are living proof of this.

This never was a technical issue, or one which could cost them money, but a cold blooded business decision which generated thousands upon thousands of kilos of e-waste.

And for what?

Looks like her husband shares that opinion.

Also the fact that they have millions to bribe (or in friendlier terms "donate") their way out of anything, and have a network of very powerful friends and allies does not help at all.

Regulation should have started when the likes of Microsoft and Google started to become omnipotent and the big players in the Valley were not as powerful and influential as they are today. Now it's probably way too late.