Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
lukev
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
46 ms
·
661.
▲
by
lukev
4y ago
Hah, I could tell because in the "feature matrix" the Hudi column was mostly green compared to the others. Immediately made me suspicious so I looked it up and sure enough, not exactly an unbiased source. Feature matrices are extr
662.
▲
by
lukev
4y ago
Update: I read a tweet thread by the author of the original post, and they also used an uHoo. So if you find the "uHoo doesn't measure NO2 anyway" argument compelling, then really we're all just talking about nothing.
663.
▲
by
lukev
4y ago
Methane (at the volumes we're taking about) does not have significant health effects; the paper you link speaks mostly about its effect as a greenhouse gas. And I'm measuring my NO2 and it's within safe levels. So should I be
664.
▲
by
lukev
4y ago
This is the same sensor used by others who have reported high NO2 levels correlated with their gas stove burning, though, so it's got to be measuring something relevant.
665.
▲
by
lukev
4y ago
That's true, but it's also valuable for analyzing air quality for lots of other things (including Covid safety using CO2 as a proxy.) I use it at home and also bring it to the office occasionally to make sure things are reasonable
666.
▲
by
lukev
4y ago
This is true for resistive electric stoves, which "pulse" on and off to achieve a certain heat level. Truly frustrating to cook with. Modern induction stoves are much better (even better than gas, I hear, though I haven't use
667.
▲
by
lukev
4y ago
Gas stoves (like everything else) have pros and cons. CO2 emissions are certainly a major con, which is why my next stove will be an induction. I got a gas ten years ago because induction cooktops were not widely available and the cook qual
668.
▲
by
lukev
4y ago
Counterpoint: I also measured the pollution from my gas stove (using the same meter the author is, from the look of the graph) and it was well within acceptable parameters. And I don't even have a vent above it. I bought a meter and st
669.
▲
by
lukev
4y ago
I do see what you are saying, though I think a lot of what they say makes somewhat more sense if you are familiar with their consulting "bubbles" and the language used within them. But the problem exists even on a concrete technic
670.
▲
by
lukev
4y ago
If you're referring to the holidays: this did not occur for the past several years. Typically they will scale up temporary staffing to handle demand. If they failed to do so for some reason, that is also interesting to me.
671.
▲
Ask HN: What's going on with Amazon Prime shipping times?
55 points
by
lukev
4y ago
|
44 comments
672.
▲
by
lukev
4y ago
There's a cost/benefit curve here. Having data can undeniably be useful. The question is, how much ROI does the insights in the data provide, relative to the cost to build the organizational infrastructure capable of extracting th
673.
▲
by
lukev
4y ago
I have no doubt the problem is worse in academia, but to be quite honest nobody is really doing a good job at curating, governing and unifying data across large organizations (at least without extreme expense.) Just look at the treadmill
674.
▲
by
lukev
4y ago
There are definitely mechanisms to do this in the JVM, such as the `finalize` method or in more recent versions, the `java.lang.ref.Cleaner` class. However, using them correctly in the context of memory leakage and safety is non-intuitive a
675.
▲
by
lukev
4y ago
Infrastructure automation tools like Pulumi or AWS CDK make this relatively straightforward... just deploy a new stack (or use the APIs to automate a deploy/test/destroy cycle.) It's even possible with good old Terraform, use
676.
▲
by
lukev
4y ago
Musk has engaged in a ton of tribal signaling recently. Does it suck that we have such clearly delineated tribes? Yes. But Musk is also smart enough to realize that that's the case, and he's knowingly chosen to repeat and engage w
677.
▲
by
lukev
4y ago
A very large part of Tesla's value was never based on any kind of fundamentals, but on optimism about Musk's vision and ability to innovate into the future. I wonder how much of the drop is based on perceived lack of "focus&q
678.
▲
by
lukev
4y ago
Solving security issues and maintaining secure systems isn't your problem, but adhering to company policy most certainly is, and the policy almost certainly forbids you from sharing passwords (especially via Slack.) Unless you're
679.
▲
by
lukev
4y ago
This is interesting. I read this link because I've been curious about exactly how these work, and I like knowing that sort of thing. Turns out, it works precisely how I would have expected it would work, and indeed, probably using the
680.
▲
by
lukev
4y ago
This is a dangerous rhetorical maneuver, though, because it: 1. Accepts the existence and influence of social-media-induced illness: 2. Does not rigorously define the condition, it's symptoms, scope or mechanism: 3. Therefore, leaving
681.
▲
by
lukev
4y ago
Same as we do now. Nobody currently goes around reading random books or articles. We use link aggregators from communities we trust, reviews from reputable establishments, or recommendations from our friends. These mechanisms are certainly
682.
▲
by
lukev
4y ago
The answer is not to "disconnect". The answer is to read critically, to read for argument, and to try to connect with the intent of what the author was saying. All the while realizing that most of the content out there is bullsh
683.
▲
by
lukev
4y ago
I believe this blog is written by GPT-3.
684.
▲
by
lukev
4y ago
It's a huge failure of reporting that this article contains precisely zero data about how dangerous this practice actually is is (or, alternatively, if it results in measurable harm like increased congestion or transit times.) I'm
685.
▲
by
lukev
4y ago
> Average mutable reference usage per repository: 1.94 > Repositories with no mutable reference usages: 7,245 (63%) > Truly bananas. Clojure libraries really do have less state. It probably isn't surprising if you've been
686.
▲
by
lukev
4y ago
> For example, GPT-3 has been used to generate essays on complex topics such as the philosophy of mind, and these essays have been well-received by readers. That's the other problem with GPT-3... it will happily say things that are
687.
▲
by
lukev
4y ago
So, I actually agree with you there. Crafting a prompt, curating the output, and then distributing it certainly itself an act of communication. I don't particularly like it as an act of communication because the information density i
688.
▲
by
lukev
4y ago
This makes me sad. I hoped to see less use of language models in the wild; looks like we're heading the other direction. I was recently able to articulate to myself why I dislike language models like GPT-3 so much. Language is fundamen
689.
▲
by
lukev
4y ago
This is a radical and somewhat risky plan. But I think we should fully explore it. The best plan, Plan A, would be drastically and immediately decarbonize and cut CO2 emissions globally. However, we have ample evidence that that is not go
690.
▲
by
lukev
4y ago
Lots of discussion in this thread about the accuracy of the implied claim "vaccination can prevent Alzheimers", but not a lot about the correct course of action. Which seems to me to be, unambiguously, you should absolutely be
More ›