bt848

What would you do with a totally open specification?

You are absolutely correct, which is why the bulk of Gmail’s client code is dedicated to hiding latency.

I’m curious how they do competitive analysis. They want to be faster than gmail but how do they measure? I tried to take a high-speed video of gmail navigating from thread list to thread but it was not conclusive. Gmail collects a ton of client-side timing data and I wonder if anyone has ever just intercepted that.

It’s wild that Vox is an amateur baseball blog that mushroomed into a billion-dollar media conglomerate.

LTO doesn't suffer from those things. Shock can destroy the case and dislodge the leader pin of an LTO cartridge but AFAIK it can't damage the data on the tape. Water is bad for the transport but as long as you dry your tapes out they should still work. The operating specs for LTO are crazy, you can read and write them under all practical atmospheric conditions.

Garmin. You know, the brand that is 100x more popular among bicyclists.

Apple marketing machine in action: get local press to write up as a novelty feature that main competitor has had for ages.

You're right but it says that the contractors who unionized work for an IT outsourcing firm.

Google has kind of a random relationship with various classes of employees. There are contract cooks and janitors and whatnot but they made the physical security guards full-time.

It doesn't. Where does the USW come into the picture, if PATP is affiliate of the DPE, considering that USW is _also_ an affiliate of DPE? I don't get it and I wished the article explained it.

Considering that they work for a third party I imagine it is the third party's problem.

They stopped issuing major updates for the Nexus 5X very shortly after Google Fi stopped selling them brand new on their website.

That's true, also government programs to create jobs are pretty misguided in an environment with all-time-low unemployment. We should be trying to create employees.

Toyota doesn't even bother trying to market the Tundra outside the Americas, since cowboy cosplay isn't terribly popular abroad, so it makes sense that they are made here.

Which part of that wikipedia article strikes you as characteristic of a "long, steady decline"?

Is that your personal definition? On App Engine I deploy a function that gets invoked whenever an HTTP endpoint is requested. It clearly maps to the semantic of "function as a service" and I don't know about or control the number of servers so it also has the properties of "serverless".

Maybe but this post is about making the two sides of the computer NOT communicate.

Skylake-X has 44 PCI 3.0 lanes, that's 352GT/s or about 345gbps application bandwidth. It's certainly more than enough to push 100gbps from disk to net. These guys are pushing 200gbps, but they're doing it with two CPUs, two sets of NVMe devices, and two NICs, and a bunch of hacks to make the operating system pretend all this stuff is not in the same box. It seems way more straight-forward to me if they had made it all be actually NOT in the same box!

So why is it important to have a multisocket NUMA machine? Why not just save yourself a lot of hassle by having one socket? I know that the previous generation AMD machine had unavoidable NUMA but the new one doesn't.

This is neat. The obvious comparison is glog's DLOG feature.

The answer really depends on the nature of these transactions. For example if you can achieve the desired throughput with one thread that could change your decision.

There's always sailing, although it's a bit laborious, and fission has a good record at sea :-)

Even if Hypeloop worked, which hasn't been demonstrated, its nameplate capacity is close to zero. It can optimistically replace one highway lane. It cannot possibly displace air travel unless you have dozens of tubes. Each of the Bay Area's three airports averages about 3-4x the passengers that a hypeloop might be capable of moving, and that's averaged over the whole year. Peak day/hour air travel can never be replaced with hypeloop.

Ordinary steel-wheel high-speed rail can do it, though. There are many stations in the world that can move 100k pax/hour.

You seem to have conflated CO2 with other pollutants. All the world's shipping taken together amounts to only 3% of the carbon emission inventory. It certainly is not the case that 15 ships emit more carbon than all cars.

When you deploy a datacenter at this scale you contract for an amount of power and you're basically obligated to use it. The trick is to get more compute out of the same amount of power.

By the way if you still have AC UPS systems that is probably your #1 problem and you definitely will not benefit from some fancy machine learning thing. Just get rid of most of your AC systems. For example you may benefit by adopting Open Rack 48V-to-point-of-load scheme that uses an in-rack DC UPS.

https://www.opencompute.org/files/External-2018-OCP-Summit-G...

Ehrlich's timing was not great and his book is almost unreadable but the things predicted therein have come to pass. It really is more crowded in Yosemite than in downtown LA, and the air quality is worse. There really aren't any non-toxic fish left to catch.

Just like basically this biggest nit ever, but Snowden revelations didn't reveal anything and nothing was described in them. Snowden exfiltrated a bunch of powerpoints that neither he nor you nor I can understand out of context. The most supportable conclusion you can really glean from Snowden's slide decks is the US intelligence agencies were able to intercept traffic of undersea cables.

Google’s datacenter power management relies on software features that nobody outside Google is prepared to accept. Most operators expect to have thermal headroom for short CPU clock rate bursts and they don’t want some central daemon downclocking their machines. From what I’ve seen outside Google, most organizations would get much more benefit from focusing on utilization than trying to replicate Google fan speed control techniques. If you still have UPSs and dual AC-DC power supplies in every box and your PUE is 2.0 you don’t need the thing that allowed Google to go from 1.11 to 1.10.

It is not possible to control rodents by predation.

Curious how many times you and millions of others thought you could slaughter all those insects with your cars before it would harm them. Roads are one of the main reasons for declining animal populations and insects are impacted as much as or more than any animals.

The form of "assert" is not important. It is isomorphic with if (a>b) {exit}. The compiler can assume that thereafter a<=b. until one of them is modified.