Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
aaron-santos
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
15 ms
·
91.
▲
by
aaron-santos
5y ago
The article contains a link[1] which demonstrates one such device in use. If you do not wish to hear pleading and screams of pain then I suggest you not watch it. [1] https://www.youtube.com/watch?v=YcxpGKctZMs
92.
▲
by
aaron-santos
5y ago
> The second is that it's all about negotiation. Companies by and large have the upper hand here, and would pay much lower wages if they could. Employee wages are roughly bounded below by the cost to keep them alive, and roughly bou
93.
▲
by
aaron-santos
5y ago
Disclosure: I read the slides, but have not watched the video. What the author is doing is disentangling accountability and support. The nice-kind distinction is reminiscent of the permissive-restorative distinction in the social discipline
94.
▲
by
aaron-santos
5y ago
In agreement. Average citizen preferences have no bearing[1] on the adoption of legislation. [1] https://scholar.princeton.edu/sites/default/files/mgilens/fi...
95.
▲
by
aaron-santos
5y ago
If you want to draw from the well of business insider they have another article[1] on that very topic. [1] https://www.businessinsider.com/racist-origins-marijuana-pro...
96.
▲
by
aaron-santos
5y ago
Active vs passive voice in media is a way of avoiding responsibility. It shifts the subject from those who transgressed to the instrument of transgression, or to the person wronged. What it should read is "Natalie, Facebook may have ex
97.
▲
by
aaron-santos
5y ago
You have the agency to choose your beliefs. The door is open, you simply have to walk through it.
98.
▲
by
aaron-santos
5y ago
Normative[1] economics[2] is[3] definitely[4] a[5] thing[6]. [1] https://www.investopedia.com/terms/n/normativeeconomics.asp [2] https://en.wikipedia.org/wiki/Normative_economics [3] https:&
99.
▲
by
aaron-santos
5y ago
> Making friends with employees This is a lesson I learned in the last year. Even though I've made real friends with colleagues and I wouldn't change that, I would still advise caution. It can be tempting to make friends at wo
100.
▲
by
aaron-santos
5y ago
I wish Brendan Eich's original intention to put Scheme "in the browser"[1] had succeeded. There was never any reason to invent a new browser language when perfectly good one's already existed. [1] https://book
101.
▲
by
aaron-santos
5y ago
> I think the punishment for corporal punishment should be jail and corporal punishment. Would that not entrench and perpetuate the harm that occurs with corporal punishment and in turn replicate the cycle of violence?
102.
▲
by
aaron-santos
5y ago
If the notice posted by Bill Anderson is any indication of the type of person he is to work with, he would have to pay a premium to make up for his disposition. Sometimes the explanations can be just that simple.
103.
▲
by
aaron-santos
5y ago
I'd love feedback too. I tried my hand at using scipy.optimize to tune Blender lighting and materials to a reference image that didn't turn out how I expected. Would love to hear other about other ideas in this space.
104.
▲
by
aaron-santos
5y ago
I've picked up Blender in the past few months. Previously I had dabbled with 3dsMax in the late 90's in highschool, but the piracy and license cracks eventually turned me away. Until recently Blender's UX kept me away too. It
105.
▲
Ganzflicker Experience
(docs.google.com)
1 points
by
aaron-santos
5y ago
|
0 comments
106.
▲
by
aaron-santos
5y ago
That's fine, at least you're open about the knowledge for knowledge's sake position. There's more than one way to judge something.
107.
▲
by
aaron-santos
5y ago
Thank you for calling this out. It's critically important that people understand the difference between model, code, and paper and what they mean. It's also important that people understand that even if code is provided, it's
108.
▲
by
aaron-santos
5y ago
I can see how it would be useful when talking about social reproduction theory, but I would still tend to favor PMC in those conversations.
109.
▲
by
aaron-santos
5y ago
Upvoting and monetization come to mind. They are really the go-tos when it comes to engagement and growth.
110.
▲
by
aaron-santos
5y ago
> Utopias in general are a great demonstration of man's hubris, and we have known that forever. This is a message that programming language designers and advocates need to hear. The idea that a better language can lead to better out
111.
▲
by
aaron-santos
5y ago
The conspiratorial premise is an external locus of control. Everything else is world building.
112.
▲
by
aaron-santos
5y ago
> Because the Hippocratic oaths are reserved for doctors of medicine/osteopathy , not for psychologists. Yes, that is the point.
113.
▲
by
aaron-santos
5y ago
That's fair. I don't know too much about other industries. In software, if you're using passion as a hiring indicator, you're contributing to this pattern whether you want to or not.
114.
▲
by
aaron-santos
5y ago
That's all well and good, but no one is innocent[1], so the only winning move is not to play. [1] https://mises.org/library/decriminalize-average-man
115.
▲
by
aaron-santos
5y ago
> Forensic psychology has a lot of skeletons in its closet. I'm in agreement. One minor example, for as much as anyone thinks the Hippocratic Oath has worth - let's be honest, this is HN, there is sure to be a spectrum of opini
116.
▲
by
aaron-santos
5y ago
Thank you. It's a tough habit to break after reading it for so many years.
117.
▲
by
aaron-santos
5y ago
JCS Criminal Psychology is a fantastic reminder and deep dive into what happens when you talk to police, why you shouldn't, and what you should do instead. It details interrogation techniques [Reid] in a way that made me understand pre
118.
▲
by
aaron-santos
5y ago
> The problem though, was that no-one on the team actually cared about growth. “Get people home in time for dinner” had to become something more like “get even more people home at some time”. Once you see this pattern you cannot stop see
119.
▲
by
aaron-santos
5y ago
Loads of questions that might help to find your answers: What's your re-projection strategy? Are you at liberty to apply the same projection to all of the data in your pipeline? If not (using UTM for rasters for example), what are the
120.
▲
by
aaron-santos
5y ago
It also covers the case where repeated sunburns increase ones chances of malignant cancers.
More ›