Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
sargun
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
42 ms
·
391.
▲
by
sargun
9y ago
So, dumb question. One of the biggest impingements upon grey businesses (prostitution, pornography, imitation goods, gambling, social media profiles) is payment processing. Now, VISA, AMEX, etc.. will never want to process these payments. I
392.
▲
by
sargun
9y ago
Cool! Just curious, what's the use case? Are you trying to integrate Haskell and Elixir / Erlang? From my understanding GHC produces faster code than the speed at which BEAM can interpret code.
393.
▲
by
sargun
9y ago
I think the cgroup BPF filters align more with what Little Snitch can do as opposed to the TCP tracepoints. In modern versions of systemd, you can restrict a given network application to only having access to specific IPs via this mechanism
394.
▲
by
sargun
9y ago
I think the formula is simple — if the amount of idle developer time waiting for development gear + the cost of development gear + the cost of porting your software to POWER > the potential cost saving of POWER, don’t touch it. If POWER
395.
▲
by
sargun
9y ago
If Google does a better job of defining who should be killed, and it siphons money into the Bay Area, why shouldn’t we be supportive of google’s involvement? I agree that perhaps we shouldn’t blow people up overseas, but at least while we a
396.
▲
by
sargun
9y ago
You work at Netronome? I’m curious if we will ever see your chips in traditional data planes like switches.
397.
▲
by
sargun
9y ago
Mmap’d IO is still a shit show because of clearing the CR3 register on page faults.
398.
▲
by
sargun
9y ago
Yeah, as far as I can tell, on a lot of these products volume discounts can easily be in the 95%+ range.
399.
▲
by
sargun
9y ago
Eh, if you use lxcfs with LXD, LXC, or docker, you can fake uptime.
400.
▲
by
sargun
9y ago
Have you tried to debug the same two problems when you do not have access to the infrastructure, versus when you do have infrastructure access? It's a radically different experience. In addition, even if no one needs access, a bad acto
401.
▲
by
sargun
9y ago
Eh, I think the only thing that works is you kill your velocity and institute a bunch of processes, and bureaucracy or put up light gates and hire people that you trust to do the right thing. It sounds like they have some gating, but perhap
402.
▲
by
sargun
9y ago
How is the podcast functionality? My biggest frustration with smart speakers is the poor support for pausing and resuming podcasts.
403.
▲
by
sargun
9y ago
I wish it was easier to get your hands on refined isotopes. I believe it'd spawn quite a bit of interesting industrial research that's being held back right now. A lot of this waste is a nuisance, but I believe if it was easier to
404.
▲
by
sargun
9y ago
Mill feels like the GNU Hurd of CPUs.
405.
▲
by
sargun
9y ago
Is Mill ever going to come out with real hardware that "normal" people can use? Or is it just an R&D project, and focused on selling the IP?
406.
▲
by
sargun
9y ago
Certain Atom CPUs have no speculative execution nor OoO execution.
407.
▲
by
sargun
9y ago
Does TrustZone mitigate this? Can the secure world be compromised via this side channel?
408.
▲
by
sargun
9y ago
Could be a Toughbook, although doubt it. Introduced in '96: https://en.wikipedia.org/wiki/Toughbook
409.
▲
by
sargun
9y ago
There is Buildkite, which a hybrid server side / client side solution. This has worked extraordinarily because one of the most frustrating parts of setting up a CI service is setting up a long-lived, reliable server.
410.
▲
by
sargun
9y ago
Often times devices of secure randomness run the bell test to ensure that they aren’t being tampered with. I believe that blasting the counter with a gamma ray beam would indicate this.
411.
▲
by
sargun
9y ago
Because installing a tracker on someone’s car will end with lots of privacy issues? You need a tracker to properly attribute taxes to the person that owns the road, and even though you don’t need super accurate trackers, any tracking can be
412.
▲
by
sargun
9y ago
Erlang is pretty poor when it comes to P50 latency compared to C++ and Java. Part of this is due to BEAM being a pretty unsophisticated VM, and part of this is due to the language and VM promoting inefficient patterns, (FP, immutability—the
413.
▲
by
sargun
9y ago
Independently from this, IPv6 doesn't really work in AWS VPC. For example, ELB / ALB breaks under IPv6 endpoints. A lot of VPC services aren't available on IPv6. Metadata service has no IPv6 equivalent. I'm sure someone
414.
▲
by
sargun
9y ago
Err, I think the question is a little bit more complicated than that. You can run various "Soft IP" cores on FPGAs, and then run Linux, or whatever you want on that. I believe right now, OpenRISC, and RISC-V cores are some of the
415.
▲
by
sargun
9y ago
How would you implement types? receive / send make dealing with types incredibly difficult. Also, hot code reloading makes reasoning about types difficult as well. There's been some proposals for Multi-party session types, but non
416.
▲
by
sargun
9y ago
Like what?
417.
▲
by
sargun
9y ago
I'd argue that switching kernels is pretty easy. At least for us, there's a lot less QA to switch kernels than Docker versions. The more often you upgrade, the easier it is to upgrade, as with almost all things. We also have a tea
418.
▲
by
sargun
9y ago
At Netflix, we have considerable use of the autoscaler with our custom scheduler. Is there any reason you can't autoscale based on queue depth, or time of items in queue? As far as your Docker problems go, have you tried to upgrade to
419.
▲
by
sargun
9y ago
How does this compare to the Crazyflie 2.0? (See: https://www.bitcraze.io/crazyflie-2/ ).
420.
▲
by
sargun
9y ago
Are there any guarantees about (re)use of stack addresses?
More ›