mrwilliamchang

To see memory consumption on the TPU while running on GKE you can look at kubernetes.io/node/accelerator/memory_used

https://cloud.google.com/kubernetes-engine/docs/how-to/tpus#...

This is like a psychopathic Clippy.

There is also an issue of compatibility.

Underneath the hood other serverless technologies like lambda are running lightweight VMs running linux. Therefore they can easily accept any linux compatible container and they can run it for you in a serverless way.

Cloudflare Workers are running modified Node.js runtimes. You can only run code that is compatible with their modified Node.js runtime. For Cloudflare to be able to offer a product that runs arbitrary linux compatible containers they would have to change their tech stack to start using lightweight VMs.

If you want to run Node.js, then Cloudflare Workers probably works fine. But if you want to run something else (that doesn't have a good WASM compatibility story) then Cloudflare Workers won't work for you.

Depending on your use case, there is a way to proxy gRPC to Cloud Run in a slightly hacky way leveraging the fact that outbound gRPC works.

You can run in GCE a gRPC server that whenever it gets a gRPC request, it temporarily stores the gRPC message and associates it with a session ID. It then sends a HTTP request to Cloud Run with that session ID. Then your Cloud Run instance will take that session ID to make a gRPC connection to your gRPC server in GCE. This GCE instance will then take the session ID, retrieve the gRPC request, and forward it to the Cloud Run instance.

This is admittedly hacky, but depending on your use case, may be good enough.

IPv6 TCP and UDP outbound works. :)

Can you clarify what is not working for you? IPv6 should work. (I work on Cloud Run).

"There is as yet no absolute challenger to the relational model. When people think database, they still think SQL. But if there is a true challenger, it is in the graph model."

This article is quite biased towards graph databases with regards to the SQL versus NoSQL tension. This video presents a much more balanced view of SQL versus NoSQL, in my opinion. https://www.youtube.com/watch?v=qI_g07C_Q5I

Here's my attempt at a summary. Wolfram speculates that the universe is a network of nodes and connections. This network changes over time by simple substitution rules. Specific patterns in the network give rise to effects that on the larger scale that we experience as the physical world. He has shown that if you model the universe this way, you can neatly derive special relativity and general relativity. He is also doing a brute-force search through networks to look for one that exhibits properties like our universe.

I'm curious if I got it wrong.

Google says they are working on a distributed version and will release when it is ready. https://github.com/tensorflow/tensorflow/issues/12#issuecomm...

"Paul added an assumption to his article, possibly after William read it, which is intended to rule out his posited distribution"

Yeah. That is what happened.

pg's argument is that if the average performance of one admitted group is better than the other the admission process has a bias. This example shows that you can have an unbiased process, but the average performance of the groups differs.

I think I have a simpler counterexample to disprove pg's hypothesis than any other counterexample I've read in the comments. Suppose our goal is to admit the top 5 applicants with the following performances:

  A - 30,000
  A - 10,000
  A - 9,000
  B - 7,000
  B - 5,000   # Cutoff point below this line
  A - 4
  B - 3
  B - 2

I have no first hand knowledge of this, so I'm probably wrong. Maybe it's more accurate to say that the late stage investments are driving valuations up (ie the a16z analysis). But a lot of these later stage investors would not have felt comfortable without the liquidation preferences. The key distinctions is that liquidation preferences are not critical a VC's decision to invest, but are critical to the later stage large institutional investor's decision.

This looks like a good plot twist for a movie.

I think the heavy reliance on these templates is a good clue. Apple TV's interaction model is a lot about swiping around from tappable element to tappable element. Arbitrary layouts will make this a miserable experience. Apple probably wants to establish a number of layout patterns that they believe will yield a good experience.

Myra looks similar Magic to which was launched earlier this year. Also has a name beginning with m. http://www.wired.com/2015/03/stump-magic/

The original Google Now team leaving is not indicative of a problem. Google Now is more a search product than an Android product, so it makes sense for it to be taken over by Search engineers.

I've written some more fleshed out thoughts here: https://medium.com/@mrwilliamchang/onhub-is-not-a-router-foc...

I think the best way to judge this product is as a hub for a smart home. The fact that it is an easy to wifi router is incidental (in spite of the fact that is all the marketing is talking about). It seems pretty clear, "onHub" is suppose to be a hub for a bunch of "on" branded smart home devices.

Sounds like Alphabet is setup to do lots of major acquisitions. Game changer for tech ecosystem.

Made me see the world a little differently. “The Chinese will sell people anything they like. They don’t ask any questions. They don’t care what you do with what they sell you. They won’t ask whether the Egyptians are going to hold elections, or repress people, or throw journalists into jail. They don’t care. The Americans think, If everybody is like me, they’re less likely to attack me. The Chinese don’t think like that. They don’t try to make the world be like them. Their strategy is to make economic linkages, so if you break these economic linkages it’s going to hurt you as much as it hurts them.”

Great quote towards the end of the article: "Not all students are convinced they need to know a great deal about U.S. history anymore. Some​ believe in the power of the global marketplace to shape their present and future lives, and therefore see our hallmark U.S. institutions — the Constitution, citizenship, federal government system ... ​and the histories attached to them — ​as arcane compared to the new worlds that technology, innovation and consumption are spawning."

Is there a way to see more of the data? I was curious to see what their data was for tech, but I couldn't find it.

Correlation is not equal to causation. The article feels like link bait to drive more attention to redfin.com.

Audio lag is definitely not the worst of Android's lagginess problems.

Love the title.

A more common mistake I notice people making is writing code that makes more memory allocations than necessary.

  # Bad: Makes an extra instantiation of a list with 1 in
  # it which then needs to be read 
  x = set([1])

  # Good
  x = set()
  x.add(1)

Great interview.

My favorite part:

Turnaround Artist Manual – Chapter 1: Walk in with a frown; blame your predecessor; slash projects, budgets, people; lower expectations, loudly; and write off assets.

I have found the Big Nerd Ranch iOS programming book to be very good. http://www.amazon.com/iOS-Programming-Ranch-Guide-Guides/dp/...

In relation to the criteria you raised the book is a good fit.

- I understand most of the basic concepts, but cannot implement them via code

I think this is particularly true with iOS programming. The book is more written in a tutorial / HowTo style (i.e. the book gives a step by step walkthrough to create an example program). Most importantly it specifies how to use the IDE to do GUI building.

- I am a person that does really well when I have a set path

Me too. Each chapter is an exercise designed to teach you some part of the iOS library. There about 30 chapters in the book and each chapter can be completed in about 2 hours.

"the screen size of a tablet is the one spec that customers clearly recognize and associate with value"

I don't think people want to pay more money for a larger screen when they are buying a tablet. Consumers have an ideal screen size (probably around 7") and a larger screen just means a heavier device. So a larger screen is more a liability than an asset.