• 0 Posts
  • 28 Comments
Joined 1 year ago
cake
Cake day: June 15th, 2023

help-circle







  • What’s more depressing is that if you make $8 an hour, you essentially get all of your paid taxes back at the end of the year because you don’t make enough to pay them. The standard deduction is $13,850 leaving you with $2790 of taxable income for the year. So you’d owe $279 in federal income tax at the end of the year, but you would almost certainly get that reduced to zero after income, child, or housing credits.

    Depressing.


  • All records created by customers, yes. The alternative used was our default structure, but that structure is also in ram. The problem faced is that a resolver can be targeted in several ways, one of which is querying for zone data that doesn’t exist. When latency is the primary performance concern, and you need a low performance cost method to look up whether a request can be (maybe) served or (absolutely) not, bloom filters look like the right filter to use.

    Performance wise (latency), there was no improvement in lookup performance, nor a reduction on CPU consumed. So for us in that application, it didn’t make sense.

    Big table was the example I was thinking of. A massive k:v store that can start at 4, 8, 32TB you can have in RAM. The problem is, you can only answer negative membership in the key space, not positive. So if you’re looking up a key you either get a absolute no or an opportunity to scan the key space anyway.

    What I’ve tended to find is that indexes make more sense in most scenarios. Not all, because there’s always exceptions. But indexes tend to be faster, still consume vastly less space that the data being indexed, and they tend to offer more powerful features rather than just membership in a set.


  • A bloom filter tells you that an element is not in a set, or maybe is in a set. The only guaranteed accurate result is not in set. You also have to reproduce the filter whenever you remove items from the set being tested, which isn’t a free operation.

    I’ve seen bloom filters abused in a lot of ways that make no sense, but very few ways that do. The classical example of scanning a k:v space in a memory efficient manner is the best application I’ve seen, but it still doesn’t really work if you need to know a key is definitely in that space. You still have to perform the lookup.

    An application for bloom filter I was working on was mitigating the effects of abusive lookups for non-existent names on a DNS platform, and it turned out that even with absurdly high lookup rates, adding an initial check for negative presence in a record set, we didn’t benefit at all.




  • Dr. Dabbles@lemmy.worldtoMemes@lemmy.mlCan't ticket what you can't catch
    link
    fedilink
    English
    arrow-up
    2
    arrow-down
    1
    ·
    10 months ago

    It absolutely does. Airplanes were being shot down in Vietnam with an SKS firing 7.62 rounds. Now, the question at that point would be do you want someone lobbing bullets into the sky so they can fall back down and potentially hurt someone or damage something? Absolutely not. But it’s not only possible, it has been done before.








  • You likely aren’t going to get enough energy to make up for the losses incurred when boosting voltage to 4.2 volts or whatever your battery requires. There’s tons and tons of scam devices out there in the world that attempt to convince people these devices make sense, but they really aren’t usable for anything meaningful.

    Charging a battery with a couple microamps per hour. Would probably negates things like self-discharge? But certainly wouldn’t recharge a battery that you have in use with a device. And if that device has radio or storage attached to it, you definitely aren’t gaining enough electricity.