It’ll take me a while to respond to this, but it is hilarious you would refer to a voice for decorum and patience as “bully class” in an article/discussion about organizing obstructionists for maximum appeal.
It’ll take me a while to respond to this, but it is hilarious you would refer to a voice for decorum and patience as “bully class” in an article/discussion about organizing obstructionists for maximum appeal.
Good point about slave revolts actually. That didn’t cross my mind. Voting wouldn’t have helped much on that front.
As for the morally justified angle, that is highly subjective. Your ideals may not align with mine. Does that mean I need to counter-obstruct obstructism I disagree with? That sounds like rapid escalation.
I did read that article, reflecting on recent perspectives, and as it is written for modern times, it raised concern; to have a outline/playbook to organize obstructionists in this climate is woefully tactless when masses are so easily enraged.
That said, there are many ways to get your message out. Websites, pamphlets, signs, heck we are Ad ridden everywhere. There is no excuse. Changing laws isn’t glamorous, isn’t fast, and isn’t easy. But the right way has no shortcuts.
Second, (in the USA) your rights end when they infrige on anothers’. To impose my needs selfishly at the expense of yours is not only infringing your rights, but possibly accruing damages.
This is not victimless behavior regardless of cause. It absolutely should not be encouraged.
It’s easy. Don’t do it.
I support protesting and free speech. I do not support disruption. In fact, it shouldn’t even be called a “protest” in the first place. Adding that tries to legitimize theft. (Yes, denied use is legally theft.) You don’t always get your way in democracy and throwing a tantrum doesn’t work.
What does is voting and education. It’s harder, but the right thing to do.
Straight up, Firefox isn’t search, so that’s never going to be competitive. Changing from Google is easy though. That aside though …
Comparing Firefox to Chrome is a little complicated as it comes default on pretty much all Android phones. Yes, we can change, but it’s still installed and running services in the background if I recall. I really hope the move away from useful extensions takes a toll on chrome and brings users over to Firefox.
Million dollar salaries are excessive IMHO and rarely justified. I’m with you on that.
Some things Mozilla does, and doesn’t do, have been instrumental in not only bringing awareness, but security for the web and triggering dialogue. That openness is important and not something Google has been known for.
Google may be covering their butt funding Firefox, but an Internet without Firefox may look much different today.
Haha guess so. Misleading graphic taken out of context. That makes more sense. I never tried the excessive water method. It sounded interesting, but I refer to my point #1 above.
Yeah, took me a while to get through that paper. Pretty detailed, but a little red flag kept waving when all the results are percentages. I finally got down to the meat of the science and was pretty confused on actual levels. I still don’t have a good grasp of exactly how much iAs they measured, as the findings are adjusted for relative exposure by body weight. I suppose that’s the point of the paper, but would have been nice to have some relatable baseline to put things in perspective.
There is also a pretty large margin of error as a exposure 2.75x greater than used to calculate the results was within the same exposure limits linked to the 0.1% risk. That’s easy to add to their charts. Would have been nice.
Wow, lot to unpack here.
“Ain’t nobody got time for that.”
How much arsenic are we talking about here?
“Saving time, water, and energy” is apparently short for boil water, stand around, drain, re-boil more water using more energy, and finish cooking rice. Where is all this time saving happening?
“The margin of exposure [to arsenic] is increased to desired levels”. Hmmm I guess if you really think about it that is a true statement, but one heck of a roundabout way to say it.
One of my favorite homebrews was made with hojicha. I’ve made it a few times.
Yup! Even units with one exhaust need 2 fans. Many standalone models even have a small 3rd fan near the condensate reservoir to re-introduce moisture, collected when chilling air, back into the exhaust so they don’t have to be drained as much.
The nice thing about inverter units is they only run the pump and fans as hard as it needs to obtain its objective. Once there, it can spin down and get much quieter. I used to have floor units and hated sleeping with those. All through the night you would get fan noise and a huge CLUNK as the compressor kicks on or off. We upgraded this year and the ramp-up is so much easier to sleep through. There is still fan noise, but much quieter and becomes white noise after a while.
While it does pull in hot air from outside, it is not cooled, but rather heated. AC units are divided with a cool side that lowers temp in a room and a hot side that extracts heat.
Warm air is used by the condenser (a radiator that collects the heat from the room) and sent back outside hotter than it sucked in.
The air on the cold side of the AC unit is recirculated into the room.
Doesn’t really apply to your setup, but I just got a u-shaped window AC from midea and it works well with home assistant. I didn’t expect it, but it also has an outside temp sensor I can read. I’ve been graphing outside temp all summer. Purple is the AC, blue is openweathermap. Ignore humidity, that is for a tropical/moss tank.
They have an indoor unit with the same wifi capability, linky so that one might work.
EDIT: you’re right, Midea compatibility isn’t standard in home assistant, nor advertised , but I added it with a custom integration. It’s been working great this summer.
And walking down to the food court to use a bathroom may be, ahem, problematic for a retirement community.
There was a dead mall in a nearby city that was finally bulldozed to make way for apartments. It’s taken decades and nowhere near habitable yet. Sort of a start though.
Kif! We have a conundrum!
Shhh! Nobody tell them about “inside out.”
Yeah they blocked it a few months ago. I don’t remember the specifics, but it seemed like a new mod/admin came into play and didn’t get the memo about Lemmy and trying to be more transparent and engaged with communities. Just poof
I find Home assistant has all these points with the exception of accessibility; I only have one account so I never encountered this scenario.
The trace feature is amazing to visualize the logic.
I do wish it had a test run feature so you can simulate different examples and data to see is the automation behaves as expected. I had one finicky automation that would fudge up after midnight.
I know datacenters have an IP range, and they could ban all IPs from all datacenters. But why?
Actually, that sort of makes sense. Reddit may have locked down content behind a paywall for profit. Whitelisting only data centers that pay would be “protecting” that.
They want unique identifiable traffic to maximize valuable data to sell. Everything you don’t want.
Oh, to me it sounded like you could access the video, it just had DRM so you couldn’t save it.
Different problem
I assume you don’t need the absolute best in quality, but I’d try a screen capture utility.
On a Mac, QuickTime player can record sections.
I’d be more disappointed in the missed naming opportunity: Upperware