Linux Mint Cinnamon is a good choice. Even as a sysadmin and DevOps engineer I use it on my workstation because it Just Works. It has good window management, settings management, file management and just stays out of the way. Flatpak is well integrated for things you may need that aren’t natively packaged, like discord.
I’ve heard good things about PopOs too but haven’t tried it.
The RAID on your motherboard is a mess and you should avoid it like the plague. — Wendell from Level1Tech
Creating RAID with either zfs or btrfs is much more easier and they perform better than motherboard’s RAID implementations. If you want a UI, you can even install TrueNAS Core as a server and manage zfs pools, share on network etc.
I’m 40 y/o, I used Photoshop & Illustrator since I was 8 years old. When I moved to Linux I tried everything, and ended up using Photopea.com and Inkscape.
I was pretty much finished writing this post until I realized you might be mistaken with how updating packages works - editing the package version field merely changes what Guix thinks the version is, not the actual package version. By modifying the version field, the source code that’s downloaded will change since the download url is conveniently built off the version variable, but the hash - and potentially the build process itself - will also change because of this. You’ll need to additionally update the hash, at the very least.
However, there’s also a comment in the definition stating “Later versions have dependencies on npm packages not yet in Guix”, so unless this comment is outdated, you’ll have to package newer versions of the dependencies too. While I believe that learning Guix packaging has been a very much worthwhile experience, you might want to use something like the flatpak Justin linked if you don’t want to go through the trouble of figuring this out right now, because as far as I can see this will not be as straightforward as just changing a version number.
Of course, I don’t have context on what you read and I didn’t look at the package definition in depth, so in case I’m the mistaken one here or you still want to know how to proceed for future reference, here’s my original post:
The easiest way to do this would probably be to use the command guix package --install-from-file=path/to/file with a file that returns the modified package.
Notably, you’ll want to also include the original define-module expression at the top to pull in necessary code, as well as add an anki at the very bottom which indicates that the file will return the anki definition:
The above method should work just fine, but I’d only recommend it for short-term usage since it doesn’t scale well nor does it take advantage of the declarative-ness of Guix.
Alternatively, if you’re looking for a more long-term solution, I would suggest either creating your own channel or setting a custom load path where you can write whatever extra code to include in your configuration. The former is the most ideal, but the latter is much easier to set up, only requiring tweaking the module name and setting an environment variable.
Personally, a channel is overkill, so what I do is globally set the GUIX_PACKAGE_PATH environment variable to my config location where I’ve defined custom modules, which I can then pull into my Guix Home configuration (including modified packages). Feel free to have a look at my config for reference, although it’s still fairly work-in-progress right now: github.com/aurtzy/guix-config
If you haven’t heard of David Wilson (a.k.a. System Crafters), he’s a great resource for learning Guix stuff, and has his own Guix Home configuration that you can check out as well: github.com/daviwil/dotfiles/tree/guix-home
I agree completely with this. At my office, I’ve started installing Krita in place of photoshop for people who need to edit images. It has its own learning curve, but it’s been a wonderful alternative.
For basic stuff it is a great alternative with a better UI than GIMP, although when you start needing more advanced stuff anything else is better than Krita for photo manipulation.
Gimp is just… not great. It’s ten years behind the times. These days I tend to use Krita, even though it’s more geared towards digital painting than general image editing.
Your motherboard supports a specific form of software raid that it expects windows to respect (with a driver iirc).
Basically that bios setting will enable it to “boot from the software raid”, then once you get to the operating system it needs to understand how to handle the software raid and make it a proper volume.
This made sense in the before-times with windows and booting from spinning disks (well, not really but still), but now you’re better off making a proper software raid yourself.
Your chipset doesn’t have a proper hardware raid controller (actually it probably has most of one by now, just not the iop itself, basically like a small cpu core that figures out what to do next), so it’s not proper hardware raid, just software games to let your cpu do software raid.
If you’re running Linux just use lvm, it’s fast and you’ll be pretty happy with the results.
Nothing wrong with hardware raid in general. But most consumer motherboards do not have true hardware raid - but instead fake raid. Which is some basic hardware boot time support for software raid. IE the BIOS can understand the basic raid features to boot the system - before handing it off to the OS to manage.
I would not use fake raid on a Linux system if you can avoid it, full software raid is just better than most consumer hardware fake raid support.
True hardware raid generally requires a separate expensive card that has its own controller and ram buffer.
Yeah, never hardware raid. It’s a disaster waiting to happen, even with expensive dedicated cards.
Easiest is with btrfs raid1, the drives don’t even have to be the same size like in other raid systems. For example you can combine two 2TB drives with a single 4TB drive into one 4TB Raid1 and also remove and change things as you want. ZFS has a few more features but is much more rigid and likely to break on Linux kernel updates as it isn’t part of the kernel like btrfs.
You can the same way you can technically run MS Office on Linux. You can manage to get it to start and run but using it for any form of productive use is probably gonna be rough.
I hate how weird the focus on arch installation is. I got attacked on reddit for calling endeavour arch. Like, I used that shitty method of installation in 1997, it’s tedious and there are better ways now.
Yeah people definitely go overboard with that. I think the only real problem is that Arch-based distros might be using other repositories and not be completely sync’d with Arch itself so then users will go to Arch support forums or communities with problems that don’t affect vanilla arch. But I’ve never really cared when people using derivatives claim they’re on an Arch system.
Installing isn’t even as big of an undertaking these days too, so it’s less of an achievement then people would like to think. Last time I did a clean install on my laptop I just ran the ‘arch-install’ script included on the vanilla ISO and it was super easy, lets you choose where you want it installed, pick from a list of desktop environments, and you pick between alternatives for other common packages then you’re good to go without having to do much manual work during the install itself.
linux
Hot
This magazine is from a federated server and may be incomplete. Browse more on the original instance.