Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
perbu
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
23 ms
·
241.
▲
by
perbu
9y ago
There is no opposition here. You can, unless I'm missing something, run a unikernel within a SGX enclave. They are complementary. Since a unikernel provides a self-contained system it could provide very clear and reasonably secure inte
242.
▲
by
perbu
9y ago
I think few of us like OCaml for its syntactical beauty. People are drawn to MirageOS for its semantics. Type safety, etc. However, there are other Unikernels. IncludeOS uses modern C++ so the Hello World looks like this: #include <os>
243.
▲
by
perbu
9y ago
Most people consider unikernels to be single process without threads. IncludeOS is rather liberal and has rudimentary multicore support. There is no preemption however so the "threads" will have to yield if there are more threads
244.
▲
by
perbu
9y ago
Wrt to reliability of thrusters - this is pretty well covered. They have been drilling for oil with unmoored platforms for decades. Here, the requirements for keeping the rig in a fixed position are quite loose compared to a drilling rig. T
245.
▲
by
perbu
9y ago
I've seen the Tesla Model X can tow horses, camping trailers, cars and boats without breaking a sweat. The power consumption naturally goes up and towing something heavy for hundreds of kilometers is problematic if you're going ou
246.
▲
by
perbu
9y ago
If it wasn't for the negative connotations people have with DOS I'd say it a lot more. Implementing a unikernel chain loader and a simple shell would be quite simple. That would be quite DOS-like. IncludeOS actually came out of a
247.
▲
by
perbu
9y ago
C++ is strongly typed. I believe there are frameworks that will give you denotational semantics in C++. It shouldn't surprise anyone that feature X is available in C++. But I acknowledge there are tons of shotguns strewn around everywh
248.
▲
by
perbu
9y ago
IncludeOS has ASLR. Or as close to ASLR as you can get without dynamic linking. We randomize the layout when we link. And since we typically re-link the image on each configuration change in a network of 1000s VMs each one will have a diffe
249.
▲
by
perbu
9y ago
Hi author here. I agree wholeheartedly. I wasn't thinking in absolutes, but I absolutely can understand how it could read that way. I could have titled the post "Unikernels are resonably secure", but then nobody would have re
250.
▲
by
perbu
9y ago
Candy Crush is compiled to more or less architecture-independent bytecode on Android and a dual (32/64bit) binary on iOS. So no worries wrt Candy Crush.
251.
▲
by
perbu
9y ago
Not sure if you're being sarcastic or not. But the Z80 is a 8 bit CPU. Linux was never ported to any 8 or 16 bit architectures - for rather obvious reasons.
252.
▲
by
perbu
9y ago
Also snow can accumulate on the top of the iceberg.
253.
▲
by
perbu
9y ago
If you look at Hiphop vs. HHVM there is certainly a significant performance upside in taking advantage of clangs or gccs ability to produce efficient code. I'm not sure why Facebook decided to stop maintaining hiphop. Perhaps manageabi
254.
▲
by
perbu
9y ago
This would be pretty useful. Something like hiphop which could transpile into C or C++ would mean we (IncludeOS) could build tooling to run a node application on directly in a vm without a underlying OS.
255.
▲
by
perbu
10y ago
Or just use IncludeOS, and you'll have the option to recompile most of the software out there (as IncludeOS is C++).
256.
▲
by
perbu
10y ago
-o hard,intr is still in my fingers almost ten years after I quit being a sysadmin.
257.
▲
by
perbu
10y ago
That has been considered. But you get into a lot of corner cases. For instance. I might suddenly need to go on a long drive. I only have 25% SoC and I will use the closest supercharger to charge up to 80% and be on my way. This way they add
258.
▲
by
perbu
10y ago
FWIW; in the current software you have to opt into "fleet learning". So the car will not participate in data gathering unless you actively turn it on. They still have access to the car and it's logs, however. Unless you remov
259.
▲
by
perbu
10y ago
One thing is the actual car. I think at least BMW should now know how to build a proper EV. But that still doesn't mean people will buy it. It would need a charging network that can help people out when traveling long distances. Tesla
260.
▲
by
perbu
10y ago
Sure. But that cost isn't proportional to the revenue generated of every app sold and every in-app purchase done on Android. Picking up a million or two in legal costs is the least they should do. If this lawsuit succeeds and this wo
261.
▲
by
perbu
10y ago
Did Uber accept any such preferences when taking investments? I thought only clueless entrepreneurs fell for such trickery.
262.
▲
by
perbu
10y ago
I think you're overestimating the battery capacity loss. There is a Dutch study which concludes the following: Based on 84 data points from the 85-kWh version of the Model S and six from 60-kWh cars, the study concludes that the Model
263.
▲
by
perbu
10y ago
The article states that the cost of a complete electric vehicle will lower than the average ICE car in 2017 and cheaper than the cheapest ICE car in 2030. So one can expect EVs to be able to compete fully with ICE cars in just a few years.
264.
▲
by
perbu
11y ago
Sort of. But logging to memory is more or less free in terms of resource costs. You can easily log one million lines per second to memory without your computer getting bogged down. This also removes the rather heisenbergian issue of turnin
265.
▲
by
perbu
11y ago
I once tried to grab a trout that was sort of stuck in a tight passage. The moment I grabbed it it immediately panicked and managed to break free from my grasp. I got wet. If I read this correctly it would not have reacted as violently to m
266.
▲
by
perbu
11y ago
I'm kind of dissapointed that the rest of the world didn't pick logging to a circular memory buffer like varnish does. It's proven extremely useful. You'll have a full firehose of information when you need it without sac
267.
▲
by
perbu
11y ago
We push data into shared memory. Then we have clients that can read the memory and present it. This makes it possible to log millions of lines per second with a very limited cost. This has the benefit of making logging more or less asynchro
268.
▲
by
perbu
12y ago
Where did you post them? Under what name?
269.
▲
by
perbu
12y ago
If you have enough memory you can just have Varnish store stuff in memory. However, there are quite a few petabyte datasets out there and there will be quite a few years before we can all stick those sets in memory.
270.
▲
by
perbu
12y ago
The kernel switched to LFU in 2.6 if I'm not mistaken. But, yes, we are. There are quite a few caches with separate eviction algorithms if you follow the rabbit hole all the way down.
More ›