Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
zwieback
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
28 ms
·
451.
▲
by
zwieback
4y ago
I reject the premise! Procrastination is a valid strategy in many situations, setting a task to a start of "as-late-as-possible" can be optimal in dynamic environments, both in projects as well as personal life. Much time and effo
452.
▲
by
zwieback
4y ago
yeah, impulse response would definitely be a good starting point.
453.
▲
by
zwieback
4y ago
For the Bell violin restoration I'd also be interested whether anyone would be able to detect a difference before and after. It shouldn't be too hard to teach a robot to bow a few notes and record whether it changed after he was d
454.
▲
by
zwieback
4y ago
Good question, I meant older CCD cameras with analog (RS170 or similar) output, digitized by a frame grabber in the PC. That was what we used for industrial machine vision in the 90s and early 2000s.
455.
▲
by
zwieback
4y ago
Very nice writeup. As I was reading it I realized what a long way we've come since I started in machine vision. We had analog cameras and doing things like bad pixel correction and per-pixel black offset correction were much harder, if
456.
▲
by
zwieback
4y ago
Me too, I also had a board with 8 relays and a centronics printer connector. Super easy to switch stuff on and off safely back in the day when a PC was a serious investment.
457.
▲
by
zwieback
4y ago
Agreed. He did say that it was "well put together and nicely done", though.
458.
▲
by
zwieback
4y ago
Same here, by mid 30s savings started becoming meaningful. I always interpreted the "always save some" mantra (which was a mantra as long as I can remember) as a way to form habits, not to really accumulate any large amounts of re
459.
▲
by
zwieback
4y ago
Wow, so much venom. I wonder what your cover letter/resume looks like.
460.
▲
by
zwieback
4y ago
Not sure about the first one. I work for a big company and when we hire we definitely human-read lots of resumes, which isn't to say that machine readable is not a plus. I agree that tailoring may not be worth it. I'm surprised no
461.
▲
by
zwieback
4y ago
I usually skim over it to see if I can make sense of it at a high level. If there are a lot of terms I don't understand I'll try to clarify those first. If there are lots of references I'll try to figure out if one of the ref
462.
▲
by
zwieback
4y ago
The pictures are entertaining but what's interesting is the idea the article raises: how often do we misunderstand old customs because we don't know whether the manuscript was meant as satire or allusion.
463.
▲
Battles of the Sexes: Duels Between Women and Men in 1400s Fechtbücher
(publicdomainreview.org)
2 points
by
zwieback
4y ago
|
1 comments
464.
▲
by
zwieback
4y ago
No, my point was that what I'm doing makes no sense and I suspect lots of people are in the same boat. Of course helmets help in the specific situation a helmet is for but it's hard to convince people to change their habit for som
465.
▲
by
zwieback
4y ago
This is how I broke my hip, it was so fast I couldn't even get my feet out of the pedals. I've been riding my bike for 50 years and I wear a helmet when I feel like it could potentially help but most of the time I don't. Stat
466.
▲
by
zwieback
4y ago
Thanks, the Talk page is especially interesting here.
467.
▲
by
zwieback
4y ago
Yep, new to me but the article doesn't explain much. I assume massive drinking could lead to all sorts of cancers as you're taxing every part of your body. It follows that a bit of drinking would have a small effect but is it enou
468.
▲
by
zwieback
4y ago
Do other countries already have them? Honestly, it probably wouldn't change my habits but who knows, it might get me from 8 to 7 drinks per week or so. What I find interesting is that it's so much worse for women, moralistic messa
469.
▲
by
zwieback
4y ago
Yeah, but can we turn "underestimating GPT" into "overestimating humans"? I think that's the direction I'm going.
470.
▲
by
zwieback
4y ago
Very interesting viewpoint!
471.
▲
Ask HN: Do humans overestimate ChatGPT because of our brains?
32 points
by
zwieback
4y ago
|
45 comments
472.
▲
by
zwieback
4y ago
In the past I worked on a project with a soft CPU (Altera NIOS) with some custom Verilog written for glue logic. It was nice because the NIOS is just like a generic microcontroller so it's painless to try out some different Verilog blo
473.
▲
by
zwieback
4y ago
No kidding, not happy about how my kids' tuition is used for bloated university operations.
474.
▲
by
zwieback
4y ago
Just looked at his bio: never.
475.
▲
by
zwieback
4y ago
... says the guy who has spent his entire career in academia.
476.
▲
by
zwieback
4y ago
I actually searched the page to see if anyone else is mentioning "managing up". Yes, I use my 1:1 for that as well. Remember: your manager works for you, their job is to enable you.
477.
▲
by
zwieback
4y ago
To me Nespresso tastes like actual espresso, Keurig is pretty trashy unless you get really high-end pods. Maybe that's not saying much since I've only had Nespresso branded pods that are expensive so it's an occasional luxury
478.
▲
by
zwieback
4y ago
Yes, one hundred percent. I work in the intersection of mechanical, electrical and computer engineering and the latter is a toddler compared to older engineering fields. I think there are several reasons for this: low barrier to entry, shor
479.
▲
by
zwieback
4y ago
Since I already use Outlook at work as my mail client I read RSS in there, surprisingly useful as lightweight reader, mostly without images like I prefer it. Doesn't have any discovery though.
480.
▲
by
zwieback
4y ago
I would think that most of the energy goes in between 12 and 3 o'clock and then flexes out from 3 to 6. So if the flex pushes back against your leg a good part of the energy should go back on the road.
More ›