Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
bogomipz
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
42 ms
·
211.
▲
by
bogomipz
4y ago
Thanks for the great links and background. I had a couple of questions: I'm guessing The Blue Book is "thee" book for learning Small Talk then? Are there other resources folks would recommend as well perhaps other books or on
212.
▲
by
bogomipz
4y ago
The exercise component is clear but could you elaborate on what you meant specifically about "food"? Do you mean a certain type of diet or only eating at certain times of the day or after working out?
213.
▲
by
bogomipz
4y ago
FSD was offered as an upgrade to EA(Enhanced Autopilot) in 2016.[1] The driver in the comment explained that he has a 2018 model and he was discussing his experience with FSD on a road trip that occurred in August through October of this ye
214.
▲
by
bogomipz
4y ago
I was reading through the comments on the electrek.co post and was struck by this particular comment: >"I paid $3K US, well before the FSD cost skyrocketed. While taking a cross-country drive in my 2018 Model 3 last August through O
215.
▲
by
bogomipz
4y ago
Relatedly there was a previous HN post and discussion about Delayed ACKs and TCP_NODELAY where John Nagle himself chimed in: https://news.ycombinator.com/item?id=10608356
216.
▲
by
bogomipz
4y ago
Are books and video generally low probability as attack vectors then? Are PDFs harder to exploit?
217.
▲
by
bogomipz
4y ago
Can Calibre deal with the newer Amazon DRM? I was under the impression that it could not at some point. Admittedly I haven't looked at either in years though so maybe that changed?
218.
▲
by
bogomipz
4y ago
Regarding 2, what is MPV here?
219.
▲
by
bogomipz
4y ago
>"Anecdotally, I flew Southwest just before Christmas. The network was already buckling and we had major delays, but we were lucky and made it through." Interesting. You don't say how far before Christmas you were travelin
220.
▲
by
bogomipz
4y ago
Under the "What About HyperThreading?" section the author states: >"Ironically, a large motivation behind SMT is likely the need to improve single threaded performance. Doing so involves targeting higher performance per c
221.
▲
by
bogomipz
4y ago
Reimbursing for lodging and meals by submitting a receipt at a later date is not the same thing as taking stranded customers out of the airport to a hotel and giving them food vouchers. There are plenty of people who have been stuck sleepin
222.
▲
by
bogomipz
4y ago
The poster states: >"If you had a legacy 5000 rounds or 10000 rounds, the # of rounds can't just be increased server side. Changing that would mean a new key is derived. So older users likely still have a pretty low number of p
223.
▲
by
bogomipz
4y ago
I think you and the OP might be making the same point - that if policy dictates that tax breaks on mortgage interest deduction are capped at 750K then why is it acceptable to give members of professional sports oligopolies tax breaks for hu
224.
▲
by
bogomipz
4y ago
Is it surprising that besides Ellison and Wang none of the other inner circle of FTX/Alameda has cut deals yet? There's at least 8 people that were likely aware of what was going on no? See: "Who's Who in the FTX Inner C
225.
▲
by
bogomipz
4y ago
Sorry if this is silly question but what is libera?
226.
▲
by
bogomipz
4y ago
Do you know this for a fact? I'm having trouble believing a bank would agree to loan the parents $250 million with only a 4 million dollar house as collateral. Even if the parents equity was 100% of the appraised value that's only
227.
▲
by
bogomipz
4y ago
>"Armchair thoughts-- makes you wonder what the game is here-- the prosecutors want a high enough bail that SBF can make, but not so high that he wouldn't pay the bond." What would the prosecutors cared if he made bail? Wh
228.
▲
by
bogomipz
4y ago
It didn't come from anywhere. From Bloomber: >"FTX co-founder Sam Bankman-Fried’s $250 million bail package is one of the largest in US history, but it doesn’t mean he actually has to put up that kind of money. At Bankman-Fried
229.
▲
by
bogomipz
4y ago
>"That is incorrect: DNS lookups happen in userspace, not the kernel." That's not even a nit, it's just plain wrong. While getaddrinfo is a library function it will invoke a torrent of system calls - socket, connect,
230.
▲
by
bogomipz
4y ago
The author states: >"When a pod performs a DNS lookup, the query is first sent to the local DNS resolver in the pod. This resolver uses the resolv.conf configuration file. In this file, the nodelocaldns server is set up as the defau
231.
▲
by
bogomipz
4y ago
>"Especially you who seem to be in decent health (you're not american so I assume you're not morbidly obese or other issues)." Obesity is a global epidemic.[1][2]. It's has been a global issue for a very long tim
232.
▲
by
bogomipz
4y ago
>"Different nodes will have different ideas of what is known and what is unknown so it may be learned unicast on the sending node but unknown unicast on a middle node." Yes, it's easy to forget that different hosts have di
233.
▲
by
bogomipz
4y ago
Thanks, these are good points and well-explained. You lost me here though: >"You're only thinking about broadcast MAC destinations but bridges also handle unicast MAC destinations. E.g. imagine NIC 1 (local station assigned MAC
234.
▲
by
bogomipz
4y ago
Thanks, I had another question after a second reading of your post: >"The first bit of an Ethernet frame is always one, which means there might or might not be an extra edge in the beginning depending on the idle state of the amplif
235.
▲
by
bogomipz
4y ago
Thanks, that makes perfect sense about things like VRRP and proxy arp, indeed they wouldn't work if the transceiver filtered on the destination MAC field of the Ethernet frame. I'm guessing this is exactly why this stuff is handle
236.
▲
by
bogomipz
4y ago
Wow, this is fantastic! I had a couple of questions. The author states: >"According to the 10BASE-T spec, each frame starts with a fixed synchronization sequence of 64 bits, first 62 of which are alternating ones and zeroes and the
237.
▲
by
bogomipz
4y ago
Isn't there still the thorny issue though that we don't know "why" some the ML works? See: "The Dark Secret at the Heart of AI": https://www.technologyreview.com/2017/04/11/5113&#
238.
▲
by
bogomipz
4y ago
It's worth noting that that this is the same accounting firm which earlier this year retracted 9 years of "statements of financial condition" that they had prepared for The Trump Organization.[1]. The Trump Organization was f
239.
▲
by
bogomipz
4y ago
Can someone say why the Power chips have seemed to not really gain traction outside of IBM? There was a moment a few years back where it seemed that Google might be getting behind the Power9 chips but then we heard very little outside of so
240.
▲
by
bogomipz
4y ago
Relatedly: "The Murky Path To Becoming a New York Times Best Seller": https://www.esquire.com/entertainment/books/a42189320/the-ne...
More ›