kartickv

I wonder how one would structure a big company to keep it more agile. Some thoughts:

- Salaries would be lower, so that people coming for the money wouldn't. There's be more variable pay. This will eliminate the "who cares if the product fails? I get a big paycheque anyway" attitude I've seen in a big company.

- When a product fails, each individual on that team will have something to lose: skin in the game. Maybe he won't get a bonus. Or will get paid less next year. Or gets demoted one level. Or something.

- Each VP would have the flexibility to pay his team the way he wants. The only thing he'll be held accountable for is the results. No salary band for each level, etc.

- He'll also be able to hire people he thinks are better. Maybe people who've been a founder or early employee of a startup more than people who've worked at other big bureaucratic companies.

- Engineering practices won't be centrally enforced, as they're in some big companies. If a team believes that unit testing every class, or code reviews that drag on for weeks, or <insert other practice here> have a low ROI, they'll be able to choose different practices. Maybe each (S)VP would get to define his team's practices.

These are just ideas. Maybe you can come up with better ones.

What would you to do to make a big company more like a startup?

Thanks for a reasonable disagreement, unlike the other respondent to my post.

I was saying that govt should perhaps not give subsidised loans to individuals (as in the Australian example) to study things like art because:

- the subsidised loan program needs to have a positive measurable ROI for it to continue. Otherwise, the next govt may bring it to an end, because there are hundreds of projects competing for government resources.

- in a world where poverty still exists, programs that are more bang for the buck may have the maximum positive impact on people. I live in India where millions are poor. Some amount of money should definitely be spent on arts / culture, but probably 100x more should be spent on pulling people out of poverty, and the way to do that is to identify high ROI investments in people.

- there are already other venues than Australian-type student loans for govts to subsidise arts / culture-type education, like university grants.

- I'm not sure about the ethics of giving a loan to a student for a course that may not pay back financially, but the loan will with the student for life, even if payment is delayed till his income increases. If arts are considered important to society, is it ethical to burden a student with the loan?

It was just a thought. I'm not adamant about it. I'm happy to see that we've had a good exchange of views, which is supposed to be the purpose of a forum like this.

Thanks for sending the link to the OECD report. I'll read it when I'm free to learn more about this topic.

Since it doesn't seem like either of us is going to change the other's mind, let's move on from this debate.

I see where you're coming from, but it's good to impose the burden of proof on the person making the claim. If the person making the claim can't be bothered to dig up any supporting information, neither can others. It keeps the quality of discussion higher than otherwise.

One one or two occasions, when I was about to post something, I looked for evidence, and realised I was mistaken, and didn't mislead others.

> Academics is necessary to teach people how to [...] understand and appreciate Art and Culture.

Any evidence in favor of that claim?

> If you think these skills can be obtained from reading books

I didn't say that nobody should study these fields; just that they needn't do so on taxpayer money.

Someone not studying with taxpayer money.

I'd think taxpayer money should be spent in ways that create measurable benefit to both the individual and the economy as a whole.

I'm not seeing the downside of only the rich being able to understand how Post-Impressionists like Suerat were different from Impressionists like Renoir.

Which doesn't provide the evidence the parent was looking for. Just because you take a variety of classes doesn't mean that they have a benefit. If they do, please give us the evidence.

Uber was an example. I was making the more general point that network effects are overrated and don't lead to winner-take-all: https://medium.com/@karti/the-winner-doesnt-take-all-57ac8af...

Uber was an example. I was making the more general point that network effects are overrated and don't lead to winner-take-all: https://medium.com/@karti/the-winner-doesnt-take-all-57ac8af...

Is PWM for varying the brightness of each pixel separately, or for adjusting the brightness of the entire screen? How are these two implemented generally on modern screens? Are PWM and FRC complementary or does one supersede the other? You can tell me about LCDs since you know them better.

For cellular on laptops and tablets to become standard, the networks should stop charging us extra for another device. I wish I could just be charged for each gigabyte of data without regard to which device is using it, just like Google Drive storage is available to all your devices without a per-device fee. As long as there's a separate fee, most people wouldn't want it.

Does this mean that the R, G and B values of any pixel is either 0 or 1 at any given instant? And how is it different from FRC: https://en.wikipedia.org/wiki/Frame_rate_control

Interesting perspective.

But that's often not the case: In India, Ola competes with Uber, Amazon with Flipkart, and so on. Some people say that network effects will lead to winner take all, but if I can get an Uber in 5 minutes and an Ola in 6, or an Uber for ₹80 and an Ola for ₹85, that's not a meaningful difference.

VCs also want big markets, but big markets are likely to have competition.

You're right that what happened in Iceland hurt Iceland. But I was addressing the more general point of why the Icelandic taxpayer should insure depositors from other countries.

> it’s illegal to murder someone regardless of the perpetrator or victim’s nationality

You're mixing up two things: whether the law as it exists was violated (in which case Icesave should of course be punished) and whether the law SHOULD protect foreign depositors. I'm discussing the latter.

> bank insurance applies regardless of nationality

I'm not well-versed in cross-border banking to say the least, but isn't that an opinion? One could just as well argue that Icelandic taxpayers shouldn't have to subsidise the British (as an example) public, by providing free insurance. If Britishers want insurance, it's up to their government to incur the costs, since it's the Britishers who'll be benefiting.

Can you give me a link where I can read the entire Windows 10 source code, make changes I want, and create a build, and redistribute it to people who already have a license?

I only came up with https://www.microsoft.com/en-us/sharedsource/ which has specific categories like enterprise, government security, OEMs, etc. I'm looking for something available to everyone.

> How did Iceland’s financial laws allow a local bank to take deposits so big the entire country can’t insure it.

Why not? Shouldn't Icelandic law protect Icelandic residents?

Even shared source would be great. You can see the code, you can audit it for security or privacy issues, disable parts you don't like and distribute modified versions to people who already have a licensed copy of Windows.

That won't dry up Microsoft's revenue stream, and is a great middle ground between open and closed source.

What if you use an ion drive or other propulsion that uses less fuel, and run it for years or decades? Yes, it generates little thrust, but won't it add up over long timescales?

Not the OP, but I found that PT has too much friction to create a task. Trello is much easier -- just type and press enter, then type another task and press enter, without even touching the mouse. I found that the friction of PT made me remember things in a mental TODO list, or Google Keep, or Inbox. Maybe it's not rational or "correct" but when tools have friction, we seek out simpler alternatives even if we know we're not supposed to do that.

The other problem with PT is that it promotes a bottom-up view of your project rather than a top-down view. You end up missing the forest for the trees. With Trello, I can create a list for a high-level goal like "Validate that users want the features we're building" and then list sub-tasks within that.

It's enriching to see how needs are different for an organisation that's not building software, but using software to build something else.

We can control polluting vehicles by imposing a pollution charge, or even banning them from some roads. Or by limiting them to only one lane.

I was trying to think in an idealistic or "from first principles" way as opposed to the practical / realistic approach you took. From that point of view, I'd want all roads to be straight lines to minimise travel distance.

> I do see it as an outdated language, just one that doesn’t really have a good replacement yet

Thanks for the perspective beyond the typical binary "X is good" vs "X is not good".

Pedestrians will also have a shorter walk if roads are straight, won't they? And, unlike cars, which can easily drive a kilometer or two further with no strain, pedestrians and cyclists are sensitive to this.

People are responding this way because there are contradictory reports out there. Some sources, like the one I linked to and Apple's T2 security document, say you can run Linux without mentioning that you need an external drive. Have you tried disabling security as Apple suggests and installing Linux?

According to https://www.omgubuntu.co.uk/2018/11/apple-t2-chip-cant-boot-... you just have to turn off the extra security.

Huh? You can absolutely grant software of your choosing permission to access the built in storage. How else does Windows or Linux on Mac work?

You disable Gatekeeper and thus run run any app without having an account with Apple.

Thanks, but iOS supports neither CUDA, OpenCL, Vulkan nor TBB. It's only Metal and higher-level frameworks like Core Image for GPU, and vImage for SIMD image manipulation.

> Does a nil unwrap failure not throw an exception of some kind?

No, it just kills the program, like a segfault in C++. Unlike Java, which throws an exception that you can catch and recover from.

> With regard to the last C++ template example, why can Swift not do this?

I'm no Swift typing expert either, but my understanding is that:

a) Swift generics are type-checked at definition time, not instantiation time as with C++. I believe other modern languages do it like Swift.

b) Swift doesn't promote types, so you can't subtract a float from a double. That rules out heterogenous types, like a float or a double. Even if both arguments are floats, the 0.1 on the right side of the < operator is a double, which can't be compared against a float.

c) Swift doesn't have template specialisation either, so we can't provide different implements, one for floats and one for doubles. If we could, we'd use 0.1f for the float case.