Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
ericd
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
41 ms
·
751.
▲
by
ericd
2y ago
Yeah, bitnets already shift those matmuls to integer math.
752.
▲
by
ericd
2y ago
Microcenter's only problem is that they haven't quite expanded to where I live. Going into one of those gives me back a bit of that feeling of going to a toy store as a kid, there's just so much cool and weird stuff packed in
753.
▲
by
ericd
2y ago
My wife jokes that we’ve never made it to the end of that museum, because the museum has always closed before we get to the end, and we’ve been at least a dozen times. I had no idea he cofounded it, among his other accomplishments.
754.
▲
by
ericd
2y ago
We didn’t start the fire, it was always burning, since the world’s been turning. Doesn’t mean it’s not worth trying to do good. It’s good for you, if nothing else. But you have to let go of the idea that you control more than the tiniest fr
755.
▲
by
ericd
2y ago
Yeah, the logistics around renting are definitely another issue. But that awful truck is just seared in my memory. No idea if it was representative, I haven't rented another since. Motivation to get over the activation energy hump for
756.
▲
by
ericd
2y ago
The problem is that oft repeated advice it makes those edge cases so much more painful, to the point where they might just not get done. Like, have you tried renting a Home Depot truck? The last one I rented was barely roadworthy.
757.
▲
by
ericd
2y ago
Oh very cool. Besides the unstable nature of L1, my main concern was actually mitigating the light pressure of the light being blocked, in order to not be blown earthward, but I guess that’s not really considering that these things could ma
758.
▲
by
ericd
2y ago
I wonder how much reaction mass we’d need every year to keep something stationed at a Lagrange point to block 1% of earth’s light, for combatting global warming. 1% of earth’s light would be a heck of a solar sail. Looks like 1% would be 13
759.
▲
by
ericd
2y ago
The other real difference is that the addressable market for most drugs is quite small in terms of number of customers. The vast majority of people just don’t need a particular medicine, especially the life-saving variety. The more niche, t
760.
▲
by
ericd
2y ago
Do you similarly object to the insane gross margins on software? Like in software, pretty sure the main expense is R&D.
761.
▲
by
ericd
2y ago
For most people, housing costing more means spending years more of your life working. Is a tiny risk reduction worth that? It wouldn’t be for me. The risk of additional years of commuting is probably higher.
762.
▲
by
ericd
2y ago
I'm guessing that the person you're responding to isn't a huge fan of huge expanses of tightly mowed grass in the form of golf courses, either.
763.
▲
by
ericd
2y ago
The batteries from eg Ego and Greenworks are way larger than anything Milwaukee puts out.
764.
▲
by
ericd
2y ago
Thanks for the thoughtful response. I was admittedly thinking more about their quality problems due to their spinning out big chunks of their manufacturing, and the general failures of management, rather than specifically the MAX problems a
765.
▲
by
ericd
2y ago
The thing is, this sets up a moral hazard where it’s advantageous not to know, not to keep close tabs, and tight control over your supply chain. If there were potentially harsh penalties for allowing this to happen, regardless of knowledge,
766.
▲
by
ericd
2y ago
Yeah, that sounds a lot like the beginning of this job, the initial plan review for the building permit took like 4 months, and I had to call at each stage in the pipeline to get it moving again. They’re always super friendly, but they don’
767.
▲
by
ericd
2y ago
Yep, going to any Bay Area city council meeting makes it very clear why it’s hard to develop - it’s an incredibly long process to get permission to build. You see the same developers again and again, making tweaks to their plans and seeing
768.
▲
by
ericd
2y ago
Any reason to go that route over just making an S3 bucket publicly accessible and pointing dns at it? Should cost roughly $0/mo unless you’re getting massive amounts of traffic.
769.
▲
by
ericd
2y ago
Yeah, but if the planning that’s meant to make things efficient is using up almost all the money, seems like maybe we need to rethink how we’re doing things.
770.
▲
by
ericd
2y ago
Yeah, I’m not sure why trenching is so expensive, I think PG&E was quoting something silly like millions per kilometer. I’m planning on doing like a tenth of a kilometer in a day with a $400 trencher rental pretty soon. I’m sure it’s no
771.
▲
by
ericd
2y ago
Yeah, that's fair. Many projects just let the free community edition support itself by hosting a forum for them to swap notes. And if a trial is about the level of complexity that you feel you can support, that's also ok. A lot of
772.
▲
by
ericd
2y ago
Usually if you think hard, you can find something that a casual user could progress to wanting more of. Or you can do what a lot of companies do and start everyone on a free trial of premium, and then fall back to free. Or the evil route, a
773.
▲
by
ericd
2y ago
Sounds about right, only a couple of their most basic models were available without it, and we really wanted induction with physical knobs, which is apparently a rare combination. So we had to give in on our no-wifi constraint.
774.
▲
by
ericd
2y ago
lol nice, thanks for the tip! I had been reluctant to soil my phone with the app, but now I guess I have to.
775.
▲
by
ericd
2y ago
This reminds me of my oven, which apparently has no setting to turn off its broadcasting its own wifi ssid. So, forevermore, all our computer wifi menus will have [LG_Oven]xyz as an option, and our spectrum will be slightly more congested.
776.
▲
by
ericd
2y ago
You can say that it’s their mistake, but people not using your software successfully is always a mix of blame for both parties. Yes, you can probably make more by adding more pressure than nagware did. But you’re doing it at the expense of
777.
▲
by
ericd
2y ago
Yeah, freemium with a free tier that’s useful for casual use converts me way more often than time-based trials.
778.
▲
by
ericd
2y ago
Idk, I’d pretty happily pay $0.50 to use an infrequently used utility once, if it was totally effortless. But everything wants like $20-30, or even worse, to lock me into some monthly subscription.
779.
▲
by
ericd
2y ago
The thing is, I often get called away from whatever I was trying to do, and just don’t revisit it for a few weeks, a month. I opened the software, but never actually used it. I go back to it later, whoops, free trial expired. I think x free
780.
▲
by
ericd
2y ago
lol awesome. Yeah, I’m kind of considering looking around for something more generic/open, but if it makes it too hard to do the basics, that might kill the enthusiasm well before we get anywhere near needing a camera. And I really lik
More ›