![](/static/253f0d9b/assets/icons/icon-96x96.png)
![](https://fry.gs/pictrs/image/c6832070-8625-4688-b9e5-5d519541e092.png)
Well I’m glad that the unifi APs like your setup better than they liked mine. Maybe they fixed it in the last 2 years. Either way there’s no way I’m buying anything else from them.
Well I’m glad that the unifi APs like your setup better than they liked mine. Maybe they fixed it in the last 2 years. Either way there’s no way I’m buying anything else from them.
Good luck if you don’t have a dream machine and you aren’t using 192.168.0.0/16. If the APs don’t find a dream machine they won’t get an IP from DHCP for some godforsaken reason and revert to 192.168.1.20 and won’t do anything until you configure them with ssh. Except you have to ssh on a lan that doesn’t exist which is a huge pita. This is why I have omada APs now.
I’ve been daily driving the pre-alpha since January, it’s definitely got a bit of jank, but it’s in really good shape. The alpha should be pretty usable, and I think by the beta it should be pretty much good to go.
When my dad died, no one renewed his domain, [last name].com, and some domain squatter bought it. A few years later the squatter noticed that I owned [last name].net and offered to sell it to me. I didn’t respond and I guess they figured out that an obscure last name isn’t worth anything and let it expire. I should probably buy it.
Doesn’t really help when control facilities are so short staffed that they have controllers working ridiculous amounts of mandatory overtime. Might well make the problem worse when they need even more overtime shifts to make up for increased rest. They need to hire more and pay way more.
For what it’s worth, just about every panel like this is certified to have a specific number of fasteners missing. A lot of the time there will be some other qualifiers such as not missing the leading fastener or not missing adjacent fasteners. Having a bunch in a row like this incident would probably not be ok, but I couldn’t say without the maintenence manual.
Got an HPE Aruba switch, it’s the only HP thing I’ve ever had that I like. Getting new firmware from HP was a pita though.
Unsurprisingly, police are considering the case as a possible murder — but the classless poll still questioned whether readers thought the woman had died by suicide, murder, or accident. Beneath the question, a disclaimer that the poll was part of the company’s “insights from AI” somehow made the tasteless poll even more egregious.
Here’s the part about the actual poll.
For network cables, FS.com. Their specialty is fiber optics and they have good transceivers and cables for really cheap prices and they also sell a tool to flash vendor info onto transceivers so if you have some picky proprietary box you can still use generic transceivers with it. Their copper products, DACs, regular cat6 patch cables, etc are good too. I haven’t tried their NICs or switches though.
The problem isn’t the manufacturer or the operator, it’s the middleman looking to make a profit on the the difference. In any case $800 is an absolutely ridiculous price point regardless of liability. I don’t know where the fair price point is but not even close to that. Liability isn’t the primary driver for the cost anyway, it’s difficulty of certification. Getting any part certified runs from high 5 figures to many millions of dollars and these are all extremely low volume parts. Boeing has only made around 11,000 737s since 1967. The plane I’m working with now only has around ~250 built since 2015 and is quite successful. For comparison Toyota produces about 20 cars per minute. When you need to pay back certification costs and turn even a modest profit on such low volume you need to charge a ton for each part.
To be clear I am absolutely not in support of non certified parts, it’s just a big problem in the industry and for rather obvious reasons.
The paperwork cost isn’t negligible at all. For example a company I used to work for had to replace a simple O-ring that failed. It’s an old part and quite rare these days and cost $800 to replace. You could buy a functionally equivalent (likely better) uncertified part for about 5 cents. That is why uncertified parts are such a problem, because certified ones are so incredibly expensive. Plenty of companies would love to step in and buy a few thousand O rings and sell them for $400 and a few are willing to forge a paper trail to make it happen. It’s a problem that I don’t really think will be ever totally solved without making certification too easy and potentially sacrificing safety by having bad certified parts.
The definitional boundary is where navigable airspace begins. You do own the non-navigable airspace above your property and you would have a trespassing argument if a drone entered that area without your permission. Where exactly the boundary is between navigable and non is a bit fuzzy but generally it will be at the highest object in the property eg. a treetop.
I still wouldn’t mess with the drone though, as another commenter said interfering with an aircraft of any type is a very serious crime.
If you can get a laptop with a few USB ports that can go a long way to helping with storage expansion. Try to avoid USB drives and SD cards, but attaching proper SSDs and HDDs with a USB caddy is a great option. Just don’t accidentally pass the boot drive USB controller to a vm like I did once.
For performance per dollar nothing beats used enterprise gear due to how little you can pick it up for on eBay. Now if you live somewhere where electricity isn’t stupid cheap or you don’t have a good way to mask the sound of a 1000 angry hornets, then enterprise might not be the way to go. Dell SFF PCs can make good servers. You can also go a long ways with just humble raspberry pis, get a whole bunch of them and you can use that to learn K8s too
I have a used 2016 super micro server. It was $600, has 2 18 core/36 thread cpus and 256 GB of DDR4 and 12 HDD hot swap trays. It also idles at 180 watts. Way over kill but I have cheap electricity and it’s nice being able to spin up a vm with just about any specs I could want. If I got some more normal cpus it would probably burn a good bit less power.
Cloudflare if you want one of the handful of TLDs they support, namecheap otherwise. For namecheap I still point the nameservers at Cloudflare so they can manage the site. For DDNS I use DDclient, it works, that’s about all I can or should say about a DDNS client.
Just to give some context, I have a one user instance running on a very lightweight Debian container containing only lemmy. After the 2 weeks I’ve had it up it’s at 6gb storage used. No clue how it would scale with more users federating with more communities but I could see it getting pretty big pretty fast.
Don’t know but it would be a good idea to ask your instance admin if you’re worried about it. They’re the ones that foot the bill for the server and it’s storage and the ones that would be doing the deleting whether using this tool or not.
Does that permanently delete posts? Why would you do that?
Reduce the footprint of the install. Text posts and comments are negligible but pictures chew through storage.
My bank gives me 0.02% on my savings account with them. My credit union gives me 3.94% on my checking account. I keep the minimum in the bank so I can use their other services, my CU only has ATMs near me.