Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
fargle
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
15 ms
·
121.
▲
by
fargle
2y ago
please, if your article is behind a paywall, post a version that everyone can read. https://archive.is/2P9Z2
122.
▲
by
fargle
2y ago
the artificial outrage over what is a simple beautiful photo lead me to actually go and find the original magazine and purchased it. should we ban it because someone knows the "dirty" secret of scanlines 600+? does it make some ne
123.
▲
by
fargle
2y ago
roshambo. southpark style.
124.
▲
by
fargle
2y ago
> practitioners’ general and typical inability to separate fantasy from reality. i don't think that is proven, nor true. simply begging the question. to hear you debate, it's as if you believe that generally anyone in the role
125.
▲
by
fargle
2y ago
> The proper way to do a vulnerability audit is by building and running code under test in sandboxed environment, and running each CVE-indicative sploit against it. that doesn't work if there isn't an exploit the other problem
126.
▲
by
fargle
2y ago
> "Straight’s path to paralysis started in the 1990s at the Saratoga Race Course". why is that ridiculous? seems tragic to me; not to mention the horse broke his neck as well.
127.
▲
by
fargle
2y ago
it's almost like, some kind of a disaster happened. i do hope they will be ok (i'm imagining severely blistered fingers in the check writing division)
128.
▲
by
fargle
2y ago
indeed. somewhere the world's tiniest violin must be playing a very sad song for her perhaps, and maybe i'm just spitballing here, the author of the article didn't agree with her take on things does she really think that if
129.
▲
by
fargle
2y ago
contrary to a sibling comment, your last few sentences are 100% right. military service -men are special because of their sacrifices they all make, whether it be health, freedom, separation, adversity, injury (mental or physical), or ult
130.
▲
by
fargle
2y ago
why in heck would you want to go there when you could just go to Oakland, CA https://travel.state.gov/content/travel/en/traveladvisories/... Exercise increased caution in the Dominican Republic d
131.
▲
by
fargle
2y ago
bingo. i think that is the best read on the situation.
132.
▲
by
fargle
2y ago
this is a lovely and well written article, but i have to quibble with the conclusion. i agree that "it’s not clear to me that NT is truly more advanced". i also agree with the statement "It is true that NT had more solid desi
133.
▲
by
fargle
2y ago
yosys [1] is the the open-source synthesis tool and also supports formal verification and other analysis. however, it's Verilog front-end is basically Verilog-2005 with a hodge-podge of newer features added on but it is unfortunately
134.
▲
Yosys and slang: A match made in heaven
(github.com)
5 points
by
fargle
2y ago
|
1 comments
135.
▲
by
fargle
2y ago
the stack itself is extremely solid and works perfectly with reasonable performance. the codebase is extremely stable and has not had major changes, nor significant bugs in ages. the biggest problem with lwip is the documentation. you have
136.
▲
by
fargle
2y ago
agree with that. it's just interesting as the story expands to see in the Amy case, there's definitely two sides to that one. btw. too late to edit, but i think i mixed up 'former' and 'later' in my comment. an
137.
▲
by
fargle
2y ago
due in part to the practical application of an eponymous law of nature discovered by Dr. Murphy, the carousels are designed to always turn in the direction that takes your luggage away from wherever you are standing. over 97% of the time,
138.
▲
by
fargle
2y ago
spent the 1/2 half of the article reading about "Amy". Her story is sympathetic, but so also is her families'. Then we get the the abrupt turn: "One woman had spent much of her twenties coming to terms with sexual a
139.
▲
by
fargle
2y ago
please, if you must post a paywalled article, add a link so we can read it https://archive.is/JAJnU
140.
▲
by
fargle
2y ago
i was thinking the same thing or maybe his boss or another higher judge!
141.
▲
by
fargle
2y ago
> Do some people's brains "just not like reading? yes, certainly. pretty common - outside of a disability, we can all learn to read, but it's hard and not enjoyable for many. > Do some people's brains "just no
142.
▲
by
fargle
2y ago
not untrue, but only part of the story. certainly your math skill level neither makes you "smart" or "dumb" (which really aren't opposites, either). prerequisites are (ahem) required . not having them does imply ha
143.
▲
by
fargle
2y ago
confused why coc is okay, dong is not okay.
144.
▲
by
fargle
2y ago
agree! they should be banned, but the FTC was clearly (over)-reaching here. the correct way to do it is via law. even though it is the right thing to do, it's just the wrong way to do it. blocking and de-incentivizing monopolistic beha
145.
▲
by
fargle
2y ago
this is a good point. there really used to be two major items that were part of this asymmetric leverage. one of them, retirement plans, have started to disappear, and that's a good thing. - as you mentioned health insurance (and to so
146.
▲
by
fargle
2y ago
all of that is fine, but my point and what i was responding to is that comparing the US system to "indentured servitude" is completely the opposite of a good analogy to what it is. you prefer "guarantees". that's fi
147.
▲
by
fargle
2y ago
> are employees actually at the level of indentured servants (which I would expect to be a contract between TWO parties) no, it's actually the polar opposite of that. either party is free to end the relationship at any time. there
148.
▲
by
fargle
2y ago
ooof. finding out that market forces are real must really suck when you "don't have the disposable income to spare" but bought those, now, 50% worthless tickets anyhow.
149.
▲
by
fargle
2y ago
well, i never said, nor do i believe that a human's will and intelligence is not a good thing in general . we're probably the pinnacle of evolution, around this galaxy at least. what i solidly do not believe is that "willp
150.
▲
by
fargle
2y ago
>"data suggest" isn't a strong argument not my argument, not my words, quoted from from TFA. > has that been applied to animals? yes 100% it has. > Does a dog communicate with his trainer with the depth and accuracy
More ›