cmurphycode

well, that's what the input boxes are for :) I don't know what the electric OR or gas rate is for wherever that person lives. But I think even your example of $2/liter, is a good thing for folks to internalize: the extremely high gas prices in europe, AFTER a worldwide systemic shock, at $7.57/gallon is break-even with a Prius at 56MPG at German/Italian prices of $.4/kwhr. Electricity is expensive, and at least in my state, I'm not seeing a serious commitment to doing something about it.

https://cmurphycode.com/electricity

Not Europe, but unfortunately, my state of Massachusetts has terrible electric costs for complicated reasons, so I understand what the OP is saying. I had to keep explaining this to my friends in MA - I replaced a Prius with a Nissan Leaf and my running costs are far higher.

(note that these prices are yearly averages for the state selected, but you can also fill in your own values since things change)

can you clarify how you got that result?

70,000 - 8800 (married filing jointly) - 2000 (two dependents) = $59200 taxable income, $2960 tax bill. An effective rate of 4.22%

To reach the site's stated 3.29%, this family would need to make around 32k a year.

Fwiw, it appears the average household income in Massachusetts is about 80k.

Hmm, I don't quite understand this methodology. This site lists my state, Massachusetts, as having a 3.29% individual income tax burden, but our income tax rate is a flat 5% (no brackets). I'm looking at the exemption list (https://www.mass.gov/service-details/view-massachusetts-pers...) and I can't see how it could possibly make that big of a difference.

I've often come to a post from https://utcc.utoronto.ca/~cks/space/blog/ in the middle of trying to understand some ZFS problem.

I was really taken aback by this given my previous experience with the Linux kernel git repo size. So, I just checked out Graal.

Receiving objects: 100% (981372/981372), 187.08 MiB | 6.39 MiB/s, done.

% du -sh graal 328M graal

% du -sh .git 221M .git

And that includes many binary files that were added and removed over time, but not filter-branched out of existence, including jars, pngs, pdfs :)

(https://stackoverflow.com/questions/10622179/how-to-find-ide...)

But the new Workspace plan is still different for 5+ users:

" Customers that have 5 or more End Users will receive a total amount of Google Drive storage equal to 5TB times the number of End Users, with more storage available at Google's discretion upon reasonable request to Google. "

Existing customers (under G-suite) may be grandfathered for an arbitrarily long time, in which case I would agree you don't need to worry.

That's correct. But the new, Google-involved behavior is that for 5+ users (which was previously "unlimited"):

" Customers that have 5 or more End Users will receive a total amount of Google Drive storage equal to 5TB times the number of End Users, with more storage available at Google's discretion upon reasonable request to Google. "

That is new behavior, and rubbing some people the wrong way.

I think the more curious thing is how long will they grandfather in the existing Gsuite users to the old, "unlimited" behavior. If they do that for long enough, at least it wouldn't feel like a bait-and-switch.

Have you tried using option-shift-[f1] to subdivide? or the brightness slider in the settings.

With a properly installed thermostatic mixing valve, you don't even need that. The water will always exit at a safe, constant-ish temperature regardless of the water temperature in the heater, by mixing in the cold supply.

"I don't think Houston is a city that was designed to be evacuated. It's just not feasible"

Such an interesting contrast to this frontpage post yesterday:

"The highways themselves were specifically intended to facilitate the reasonable objective of Houstonians not to get annihilated by a nuclear blast...[In] case of atomic attack on our key cities, the road net must permit quick evacuation of target areas"

(Discussion: https://news.ycombinator.com/item?id=24331698)

So one might ask, if all these highways don't even permit quick transit, what the heck are they doing?

It's also cool to try throat singing and watch the overtones change https://photos.app.goo.gl/DZxXKw6sKmxTJ9Wn9

Holy moly. That price per million btu is a lot less than the number I stumbled across. Makes it even more obvious.

I was curious about this a while ago and did some napkin math.

Natural gas is roughly $15 for 1 million BTU. There are 3412 BTU in a kwhr, so if you heated resistively, you'd need 293 kwhr to get 1 million BTU.

In my area, which I feel has pretty high electricity cost, we pay $.24 per kwhr, so that'd be $70.

Therefore, you need a 70/15 (4.666) COP for your heat pump to match natural gas by price. My understanding is that that would be an unusually high number for cold weather conditions.

This is close enough to what we do -- and it works -- but since we support multiple releases at a time, it often ends up feeling like merging is pointless. Merging a dev branch into trunk feels good, but honestly most git interactions involve cherry-picking sets of commits to old releases, or to create hotfixes (So much, in fact, that we built some tooling around it to make it easier to not screw up). So then, you start to wonder, why am I special casing this one situation and using merge to trunk, and using cherry pick everywhere else.

This is so awesome! Back in 2008 I built a projector from an old unsued laptop and a high powered metal halide overhead projector that I bought off ebay for 50 bucks. It was surprisingly good looking; I was very happy for the total investment. It was unwieldy and pretty power hungry, but none of that really mattered living in my college dorm at the time.

When the laptop screen eventually gave up (I think just the connector) I never ended up rebuilding it. Nowadays you can get a really awesome looking projector setup for so little, it's not really worth trying the same approach.

But from what I skimmed through, this is a build of way higher complexity, cost, and quality. I'm looking forward to watching this video tonight.

I'm on the hunt for my next laptop and I'd really like to have a rock solid portable that runs Linux, but my past experience with Dell indicates inferior hardware and construction to Mac.

I looked at the 2020 Macbook Air (the $~1300 model), and it seems to offer the same or better specs (depending how much you value the display) for the price. Are we at the point where we must pay a premium NOT to run OSX, or am I missing something? I don't love the direction OSX has headed, but I'm still content enough running it and using Linux remotely where needed.

The i3 you linked matched. But for the i7 model, apple claims 1.2/3.8 and I can't see a match for that.

For the i5, they claim 1.1/3.5 which could be a match for https://ark.intel.com/content/www/us/en/ark/products/197119/... with the "configurable TDP" of 1.1 instead of base 800 MHZ.

If anyone can solve this mystery, it would be great to compare to other laptops like the Surface Pro, whose i7 is "Quad-core 10th Gen Intel® Core™ i7-1065G7 Processor"(1.3 / 3.9)

Thanks for the new addition to my mnemonic collection!

perl -lane # process input by line, split into fields, print outputs with new lines, execute code

netstat -planet # show process, extended info, all socket types

ss -pimento # basically every piece of info about live connections

iostat -txyz # extended, omit first report (the since-boot one), omit devices with no activity, log datetime

And being a chef is knowing how to make it work anyway :)

Well, right. But, I think that a lot of people are frustrated because living in dense urban areas is NOT an allowed pattern of development in most cities, especially the Bay Area.

Hours in front of screen are perhaps the easiest metric to check up on workers. Unfortunately, it's also not well correlated to productivity! Lazy managing, for sure.

Personally, I'm pessimistic that any metric about productivity can be useful in the long term, but I recognize that in this reality, there WILL be metrics used to judge you. So, what's the best way to handle this? Is there anything an individual contributor can do to guide towards better metrics, or even just the perception of such?

In deduplication, we check to see whether a block has been seen before. If it has, we don't write it. If it hasn't, we write it down and record that information for next time. Checking if a block has been seen requires looking in an index, which is large and expensive to check, but recording the store can be amortized.

So, a bloom filter can tell you whether it's worth spending the expensive check to verify that the block is a duplicate, or whether you should skip that behavior and move on directly.

Same is true for malicious site detection. Google offers this feature in Chrome, but we don't want to make an expensive network request for every URL we visit. And we can't really have a database of all malicious URLs downloaded and updated to everyone's computer. Instead, they can distribute a bloom filter of malicious sites. if the bloom filter says it's malicious, we can afford to check the network for the final answer. If the bloom filter says it's not, we know we don't have to check (and that's the most common case, too!)

In my opinion, the intuition to extract is this. If you need to check membership, the set is expensive to check, and getting a definitive NO is valuable, then you can consider a bloom filter (or another probabilistic structure like a cuckoo filter). In dedup, we need to check membership, the index is big enough to not fit in RAM so it's expensive, and getting a definitive NO means we can skip the index check. In malicious site detection, we need to check membership, the answer requires a network round trip so it's expensive, and getting a definitive NO means we can just move on to loading the site.

Hmm. I don't think they claimed to have an O(1) time solution, just O(1) added space. Which, it is, but only because they're counting on the original array's underlying type having enough bits for their sign flipping. It would be as if you used a more compact type for the array elements, and then allocated another bitmap for the range of numbers.

Of course, once we start optimizing how the original array is stored, we may have exceeded the limits of this problem as a teaching exercise :)

As for time, it does seem to be O(n) to me; can you clarify why you think it's nlogn? It may not be particularly fast in practice when compared to other O(n) approaches like the bitmap, but I don't think the complexity is wrong.

Your solution is nice - it actually gives you more information (how many appearances, not just T/F >1 appearance), but it does require more additional space and isn't necessarily faster. I think the bitmap approach would be nicer if you're ok with using more space; the bitmap is essentially a very easy to find perfect hash function due to the unique input constraints.

18% of net is within range for fiction novels, but I wouldn't be surprised if a work like this was completely different.

Great graph! It doesn't take much away from your observation about the increase recently, but I do think it's fair to point out that you should click the "Adjusted for inflation" button for a fair comparison.

There's some evidence that grinding with an extremely high quality grinder will give better results for at least a few days than on-demand grind with a "bad" (i.e. less than few hundred dollar burr grinders, not to mention any blade grinder) grinder. This is one such "study" - https://prima-coffee.com/learn/article/grinder-basics/it-alw...

So, I think you're doing fine unless you want to spend more money :)

If you're brewing espresso, where the freshness of the grind is integral to the texture of the result, maybe this is less true. But anyone who's taking espresso seriously has a nice grinder anyway, I guess. (And on the opposite, I think French Press is very tolerant)

That's definitely part of it. I'm actually surprised it's working in heat pump mode at all. We have a hybrid and it will run pure-resistive if the ambient temperature is below 50. Now, it's in our basement, so even in the winter that's a reasonable cut-off - not sure if you got a specific model for being "outside".

Speaking of which, does your model not have a resistive backup? Ours has several modes:

- only use heat pump, unless not safe to do so (ambient too cold) - use heat pump normally but fall back to resistive to recover - use resistive only

We only have 2 people in the house so we are able to leave it on the first mode, but in the second mode it would, presumably, be the same performance as our old resistive-only heater.

I think that an underrated factor in the dynamic range compression is the way we listen to music. I'm not sure if it's accurate to say that the average quality of listening devices (that is, speakers/headphones) has decreased, but certainly it has become more common to hear music through bad earbuds, tinny smartphone speakers, or laptops simply because those devices are so prevalent.

In these settings, music with lots of dynamic range may legitimately not sound as good. Due to the quality of reproduction, the low-loudness parts are hard to perceive.

For me, there are certainly songs that I enjoy listening to on good headphones in a quiet room that don't have anywhere near the impact when played on laptop speakers.

You could estimate this in a couple different approaches, here's one.

Tesla Model 3: 26 kwhr / 100 mi Model S: 33 kwhr / 100 mi Model X: 39 kwhr / 100 mi

Hard to say where the balance of supercharging comes from (more expensive models more likely to have it free, but maybe less likely to use?) so let's be generous and call it 30 kwhr/100mi

I pay 24 cents per kwhr but I hope their average price is more like 10 cents (not even counting the solar...)

So each 100 miles charged would cost $3.

So 100 million dollars spent giving out electricity would generate 3.3 billion miles charged.

Back in september there were 20 million telsa miles driven per day, but a year long projection would be hard given the fast growth of model 3. let's just say 20M per day. That's about half of the year's miles driven covered with 100M.

I can't imagine that even 1/2 of the total miles driven are charged by superchargers. So I am inclined to agree with you; the free-supercharging-electricity costs are probably a small piece of the total 500M.