tfehring

40 ms·

391.
by tfehring 5y ago
Is denormalization (e.g. to star/snowflake schemas) still common or advisable? I know the conventional wisdom is that it can improve performance, but how much scale does it actually get you in practice (a factor of 2? an order of magni
392.
by tfehring 5y ago
Depends on what you consider "meaningful" I suppose...but yes, as long as the sample's representative. You can do a lot with n=100 - unless the authors cherry-picked it would be incredibly unlikely for them to find 83/10
393.
by tfehring 5y ago
You get an insurance card that you can take to any doctor that accepts Medicaid, they'll treat you for free or for a nominal (say, $5) copay and bill the state. Mental health treatment is covered, prescription drugs are covered, the on
394.
by tfehring 5y ago
It’s much less of an issue than it was a decade+ ago. Unemployed people in the US get free health coverage through Medicaid pretty much everywhere but the Deep South, albeit with a limited selection of doctors.
395.
by tfehring 5y ago
I go back and forth between Apollo and https://i.reddit.com .
396.
by tfehring 5y ago
I'm also not opposed to targeting in principle, and in fact I'd love to reach an end state where a company - whether you call it an ad company or otherwise - does 98% of my shopping for me. But I feel like we're in an uncanny
397.
by tfehring 5y ago
My impression is that the article mostly describes data analysts, not data scientists, though of course job titles are fuzzy and variable. IME data scientists pretty much universally at least attempt to work on things that directly and de
398.
by tfehring 5y ago
> If a company is worth its salt, it will offer services on top of existing public knowledge, without creating walled gardens. I contend that every profitable company leveraging public knowledge does offer services (broadly construed
399.
by tfehring 5y ago
Glad I deleted my account at the first sign of this sort of user-hostile behavior. Really a shame such a great and needed concept was ruined by poor execution.
400.
by tfehring 5y ago
This is what my Mazda does, it's perfect. The big knob does exactly what you describe, where turning the knob scrolls. There's also a small dedicated volume knob that you can push in for mute and left/right to change tracks,
401.
by tfehring 5y ago
According to [0] Tether has to file their first report today disclosing the composition of their reserves as a condition of their settlement. I haven't verified that, but if true the sell-off could be driven by insiders and/or sma
402.
by tfehring 5y ago
The average employer contribution to health insurance was around $5,500 in 2019. [0] I wouldn't think that reducing cash compensation by 20% + ~$1,100/year instead of just a flat 20% would be a huge deal for US tech workers. [0]
403.
by tfehring 5y ago
On your first point: the fact that there are false positives in the admission process doesn't imply that there are no false negatives. The dropout rate would presumably increase as they admitted more students, but I don't think it
404.
by tfehring 5y ago
I completely agree! It's well established that the impact of (B) is substantial, I just wanted to make the (somewhat tangential) point that wealthier students' empirical advantage at standardized tests is mostly not due to test pr
405.
by tfehring 5y ago
Tens of thousands of kids are Berkeley engineering smart but don't have the opportunity to attend Berkeley, either because they get turned down or because they (accurately) determine that they would likely be turned down and don't
406.
by tfehring 5y ago
For what it’s worth, test prep has little effect on scores, wealthy families use it at (at most) only a slightly higher rate than poor families, and black students use it at a slightly higher rate than white students. https://mar
407.
by tfehring 5y ago
That seems like a secondary concern and also mostly not Berkeley’s responsibility. The part that Berkeley is (or should be) responsible for is that the number of students who could get through its STEM coursework is orders of magnitude hi
408.
by tfehring 5y ago
More than 1% of US households (something like 1.5M households with a total of ~3M people) have >$10M in household net worth. The primary target is that class, irrespective of its members’ earned income, in part because they can reliably
409.
by tfehring 5y ago
It's really unfortunate that rent-seekers (may have) ruined blockchain as a technology. The fundamental problem IMO is that the most useful applications have no speculative potential, and structuring blockchain applications in a way th
410.
by tfehring 5y ago
Military law enforcement is going to have more leeway to conduct warrantless searches of the devices of people who haven’t even been accused of a crime? On what basis? I’m skeptical, but if you’re right that’s horrifying.
411.
by tfehring 6y ago
For comparison: - The deal negotiated by former governor Scott Walker was expected to cost taxpayers between $172k and $290k in government funding per job created. [0] - $80 million for 1,454 employees, the provisions of the updated deal pe
412.
by tfehring 6y ago
The author already explained upthread that she knows how to do that and has a working setup to do so. But that process is onerous because LKML’s workflow is technologically backwards. (That’s my claim, not hers.) Assuming that someone with
413.
by tfehring 6y ago
It’s hard to compare apples-to-apples, though, since the capital efficiency, payback period, and profit pattern are so different. If you have access to limitless, relatively cheap debt, then yeah, there’s absolutely no reason to build fossi
414.
by tfehring 6y ago
It’s harder when you’re salaried, but in service sector jobs the success rate from asking your manager for more or fewer hours is usually pretty decent IME. Less so if you’re asking for OT but only because that’s more expensive.
415.
by tfehring 6y ago
> It might surprise you to learn that this site and its newsletter costs about $130 USD a month to run Yes it does. I know you can never criticize someone else's infra nowadays without coming across as "I could clone StackOver
416.
by tfehring 6y ago
I’d also like to see more fines in the range of multiple years’ worth of revenue, but it wouldn’t change directors’ incentives since they still have limited downside exposure and therefore asymmetric payoffs. Even if a CEO runs a company in
417.
by tfehring 6y ago
Oh, if it's just a joke that went over my head then that totally makes sense! Though in fairness there are apparently-serious interview questions that are even easier (e.g. FizzBuzz, though I don't think anyone asks that anymore e
418.
by tfehring 6y ago
It's funny, I've heard "reverse a binary tree" as the canonical coding interview algorithm question for over a decade and I've never stopped to think about what it actually involves. But I just Googled it and it see
419.
by tfehring 6y ago
What is Polymarket doing in the insurance space? Didn't see anything along those lines on their site. I've previously encountered https://etherisc.com in passing, but it's not clear that they have a plan to addres
420.
by tfehring 6y ago
> Is Firefox worse than Chrome in anyway? IME Firefox is much more of a battery hog on MacOS, Chrome's history page is much better, and some sites (Glassdoor and Google Meet come to mind) don't work properly in Firefox. And I

More ›