• 1 Post
  • 41 Comments
Joined 2 years ago
cake
Cake day: June 14th, 2023

help-circle
  • Latest release seems to have been in March 2025. The version you get would depend on how you installed it. For Ubuntu/Kubuntu it looks like maybe they point to the Snap store? I have no idea how well that works and personally would avoid it.

    Did you do that, or, download a deb or install “printer-driver-gutenprint” via apt?

    In another comment you said you have the Canon proprietary driver. I think you would need the CUPS driver for this to work.

    I didn’t do this recently enough to remember what the process looked like. But in your situation I would probably try to uninstall everything print related, reboot, and then start with sudo apt update -y && sudo apt install printer-driver-gutenprint and see where that gets me.










  • If you have trouble with the soaking, black beans do very well with a “quick soak”.

    1. Cover them with water about twice the depth of the beans. Add about 1 teaspoon (~5 ml or 5-7 g) salt.

    2. Bring to a boil and keep it boiling for 2 minutes. Then cover and turn off the burner/hob. Let soak for 1-2 hours.

    3. Add any extra seasonings now (but nothing acidic). Then bring back to a boil and then simmer until soft. Adjust seasoning and you’re done.

    They should take much less time than cooking from dry. How long will depend on the beans. Older beans can take much longer, but most should be soft in 1 hour or so.




  • It’s significantly immediate-er with induction - particularly going from cool to hot. Boil water in 2 minutes and handles don’t get hot in the process. And since nothing is heating except the metal of the base of the pan there is no residual heat from the cooktop parts or the sides of the pan when you turn it off. The temperature drops much faster.

    I went back to gas after 5 years cooking on induction and miss it a lot. Cooking something like pasta that requires boiling a sizeable quantity of water takes 2x or 3x longer on gas, even with a very powerful burner.


  • Does this media server need to be accessible when you are away from home? Will you store personal data on it?

    Out of band management: this is a server feature that lets you access and manage the server even if the OS is down. That’s important if you may be away from home and need to fix a boot problem.

    You can simulate some of this with PiKVM (remote console access) and PDU solutions (remote power control).

    Redundant power: servers often have redundant power supplies, so that if one fails it can still function.

    You can simulate this, with short downtime, by having a replacement ready. Mini PCs make this easy by using relatively inexpensive laptop style external power bricks. But also think about the power circuit - is the server on the same breaker/fuse with something that could potentially take the circuit down while you are away?

    ECC RAM: this is about data integrity. If there is a failure in non-ECC then a bit flip could cause data corruption.

    You can’t really get this without ECC. Using a file system that has anti-corruption features can help reduce some of the risk. You probably trust your data to consumer PC hardware, so this would be no different really. It’s about risk mitigation.

    And that’s the main thing here, deciding on the use cases and prioritizing/budgeting how you mitigate risks to each.


  • Apron is a fun one. The Latin word for cloth, banner, tablecloth is mappa. It’s the same word used in “mappa mundi” meaning map of the world (we contract that to just map now). French often changed Latin m to n, so mappa became nappe, then nape.

    English borrowed nape directly for cloth and added it’s native diminutive suffix -kin to refer to a small cloth, a napkin.

    But there was also a similar diminutive in French - naperon. A cloth to keep your front clean. English borrowed that too, as napron. Then, sometime around the 15th century, “a napron” got mistaken for “an apron” since they sound identical. And that’s what we have today. (Source etymonline and others)



  • My understanding of why is that it relates to their change to a scene-referred workflow. Up to v3, darktable used a display-referred workflow like other programs. In that model the image you start with is mapped on a tone curve from the start where 0 is pure black and 1 is pure white, and the midpoint is set to midway between. This is all from the standpoint of what your display can render. The scene-referred workflow in v4 doesn’t do that. All the tones are mapped in an unbounded and uncurved way. So images look flat, but you’ve retained maximum data, so you have more to work with. The developers assume that you want control and maximum fidelity. There’s a better explanation in the intro of the documentation. This impacts everything - especially the color balance.

    One of the problems is that all of the display-referred tools remain as modules in the interface, and some are even used in the base processing, but you’re not supposed to use them. At least if you want to do things the ‘right’ way. We created a custom panel that has 90% of what we regularly use (shared UI with my partner). That plus creating some presets that work well with our cameras has made it very quick to get a satisfying output in a minute or less.

    Honestly, if you want to do minor tweaks to a RAW and mostly want what the out of camera JPEG looks like, there are much easier tools. If you at least occasionally deal with really challenging photos, or you want to get creative in the processing of some of your RAW images, darktable opens up a lot of possibilities, while being free and open source. So I think it’s worth the effort to learn. Shooting with a colorchecker helped us get the presets we wanted for a variety of shooting conditions.