sseth

Let's ask a question : How did Google defeat another giant - Microsoft - in 2 major markets where Microsoft had a head start and deep pockets? How did Chrome beat IE, and how did Android beat the several generations of Windows Mobile and Windows Phone?

Clearly Microsoft had no dearth of talent or money. So how come it could not keep Google at bay? In my view the answer lies in the fact that Google's engineering led culture meant they could follow a release cadence that is unheard of for such large systems. Chrome was releasing every 6 weeks, and Android every 3-6 months or so in the early days. Meanwhile we had 1.5 - 2 years between each Windows Phone release.

When Chrome started, IE had nearly 70% market share and the benefit of being deployed on nearly every computer on earth. Microsoft was in the mobile phone market from 2000. In 2011, IDC predicted Windows Phone would overtake the iPhone. Both Google and Microsoft were taken aback by the iPhone, but Google was able to turn on a dime, Microsoft was not. Google achievement has been to stay agile even as it grew big - in my view this remains the secret to their success.

At least i have found it limiting when I want have a mix of languages, and i am forced to have all the go packages in the root folder or live with strange package names. Hate is used a bit too often for my liking - nuisance would be more appropriate.

You are right that it is solvable. And in many other languages the package management etc have grown independently of the core language itself, from the community. So the attacks on the golang team seem a bit strange. Maybe they thought that these solutions would come from outside the core language team.

Later in the article ... "the White House later conceded that they hung the banner but still insists it had been done at the request of the crew members"

He stood in front of a banner put up by the White House saying "Mission Accomplished" and claimed victory. He did not clarify that the huge banner behind him was about the crew's mission and not Iraq. It is legitimate to say that Bush proclaimed it, and pedantic to argue he did not speak the words.

Perhaps instead of replacing the call center itself, it could eliminate long IVR sequences, gather better and more pertinent data from the customer and make the humans on either side more productive. As an example, in India you are often asked to press 1 for language X, 2 for language Y. But a smart assistant can just listen to the customer and decide this automatically. This can be a huge win by itself.

Is there an equivalent index where the icosahedron vertices are over land and not sea (at least for the most part)? I presume for marine requirements, such an index would reduce distortion over sea?

Actually both are true. Google (or earlier search engines) did work on basis of keywords.

The ability to handle full sentences well came much later in one of google's major search updates - I think Hummingbird which came out in 2013.

Google has dropped hints on how it may use Duplex in the way you say - reservations are just a way to train it. An example is to get better data on businesses in Google Maps by calling up businesses to see if they are open or closed on certain days, thus helping to improve one of their key platforms.

Kubernetes, CGroups (for containers) VP9, VP10 codecs

They are working "properly". Living on the edge of what the law allows and sometimes going over it is a feature, not a bug, unfortunately.

I was not clear. If you say vgo get B@latest in A, you will get the correct version of C. If you do not pull in the latest version of B, you would not get that.

The thing I like about this is that for many transitive dependencies such as C, i do not want the absolute latest version - i would prefer the version of C that B was tested with at time of release. I can override this of course, but this is the default behavior i like.

The key here is that Cargo ignores the lock files of the dependency, while vgo uses the go.mod files of the dependency.

So, if project A uses B, which uses 1.0.27 of C, then the lock file for A is locked to that version of C. Suppose B now releases a version that was tested with 1.0.28 of C, A will continue to be built with the older version because of the lock file, while vgo would (correctly) start using the new version because of MVS.

Most package managers ignore the lock files of the dependencies when picking up versions - they only use the lockfile of the topmost project for picking up versions.

The big difference with the MVS strategy in vgo is that the dependency list of a dependency is actually used to determine the version. If package A uses B, which was tested with v1.2 of package C, you will get v1.2 of C, even if there is a later version of C available.

In the typical package manager scenario, the dependency list of B may be pointing to an old version of C which does not even work with B - there is generally no way to ask : give me the latest version of C which was tested with B.

The version i like goes something like this :

Any problem of abstraction (in programming) can be solved by adding a layer of indirection. Any problem of performance can be solved by removing a layer of indirection.

There labs were in the market for years, using a "mix" of their own tests and also standard tests from other manufacturers. What they could not come out with was any FDA approved test of their own except for one (for herpes).

The basic issue is that drawing a pin prick of blood from the fingers results in a lot of "interstitial" fluid. The combination of a very small sample and poor quality made it hard to create reliable tests. In reality Theranos labs were mostly taking vials of blood from veins for standard tests, instead of the pin pricks they were touting. Once this became public the bottom fell out.

Edit: The worst part of what they did was that they would take smaller samples than normal and then run them diluted on standard machines . This is because the standard machines needed a larger volume. This gave patients less reliable results than they would have had a larger sample been taken in the first place. And i think this is really unforgivable, because it has real life consequences for people.

Waymo does nearly 5500 miles per disengagement, GM seems to be at about 1200 or so as per the California submissions for 2017. Of course this does not tell us the quality of the miles - it is easier to achieve a lower rate of disengagement on the highway for example, but perhaps it serves as an indicator.

All others testing in California are 300 or fewer miles per disengagement.

I think war is used more in the sense of a "mission" in such cases. People see the single-mindedness, cooperation and the spirit of sacrifice that happen during a war, and wonder if why we cant bring that to bear against the other challenges we face.

At work, we sometimes set up a "war-room" during production downtime - again the logic is the same : we need to work with single-minded focus on this one problem and all else is secondary to getting the system up again.

The "right to be forgotten" is always going to be fraught. We are talking of a balance between competing rights of freedom of speech and privacy, and usually such finely balanced decisions are what the judicial system is for.

But large platforms like Google and Facebook have become like "countries" in themselves. People are earning their livelihoods, finding their friends and loved ones, even committing crimes using these platforms. I think these companies are being forced to build due process procedures so people have recourse, not just for right to be forgotten, but also for other cases. And this is being compounded by the use of AI algorithms which work in mysterious ways even to their creators - an impersonal algorithm making decisions which may destroy someone's life, or make someone a millionaire.

I wonder for example if the recent youtube attack could have been avoided if the shooter had an independent forum where she could ask for explanation and perhaps challenge the change in the ads policy which hurt her livelihood. Nothing justifies violence, but we can learn lessons for the future.

We normally do not ask corporations to do such things, but for these mega-internet giants maybe the normal rules are not enough. Perhaps as a start, these companies should setup an ombudsman who can be approached when normal corporate avenues fail.

I am curious about the "cryptographic hashing technique" being proposed. How does that work? Is it just a hash of the name / dob / other identifying info? Does it somehow include matching faces?

I remember reading that when DeepMind was acquired by Google in 2014, they made it a precondition that an ethics committee be set up to review use of AI, and Google did back out of some projects because of it.

In this case what they seem to be doing is allowing the military to use some of the AI tools on Google cloud which are also available for general use. This is a much grayer area - it is one thing to not develop something specifically for the military, it is another to say that we will not provide the military something we provide to the general public.

Sorry for being out of the loop - it was night time in India, and i had gone to sleep.

You are right that the title says 2003 - i looked at the scale below which ended at 1998. I was not so much doubling down, but rather talking of a slightly different point that the ones you and jpollock were making. When I said democracies were in retreat, there are two or three different phenomena.

First - the western democracies seem to be internally weaker, and unable to arrive at consensus even for important national priorities. This may be because of the rise of social media echo chambers. In Europe there is a growth of right-wing parties in all countries, even Germany. As an Indian this worries me, because for all their warts, the liberal democracies of US and the European countries are the model of governance I would like to have succeed across the world.

Second - there seems to be no answer to the salami-slicing tactics of the autocratic regimes in Africa and Asia. You have to see the bullying of countries like Mauritius, Sri Lanka, Phillipines, even Vietnam and wonder. As the western democracies look to be in retreat, countries are surrendering their interests to the regional bullies. This will ultimately hurt all democracies in the long run.

Thirdly - there is the increasingly blatant and surprisingly successful interference in open societies from the closed ones. The US election is only the most advertised one. But similar interference has happened in Australia, in UK, in various part of Europe. There are at last signs of response to this - Australia has passed a law against foreign interference, and the Skripal episode has led to (some) concerted effort. But the response still seems much weaker than the provocation.

In summary, by "retreat of democracies" my focus was on their will and ability to stand up to autocratic forces, not a retreat of democracy itself (which i think is what jpollock and your responses were about). That may be why we were talking at cross purposes. I fervently wish i am wrong - and I hope i never try to avoid rational discourse.

Nukes are an instrument of last resort, but you don't want to get to the point where that is your only option. And it is possible that technical advances in missile defence may blunt this advantage.

Being able to prevent friendly (and even neutral) countries from being bullied or going over to the dark side is also important because ultimately the best defence against rising autocratic forces is to ensure a broad alliance of friendly, democratic countries who are secure against interference by unfriendly regimes.

If a country like the USA can have its elections manipulated, how vulnerable might smaller democracies be?

As an Indian, I hope it is the USA. At the end of the day, it has to be a multi-ethnic, multi-cultural democracy or I fear for what the world will become. And I say that even while agreeing that many military actions the US has taken in the past have been terrible.

Note : I do not agree with the petition, but the question was too tempting.

Well "governments" already have nuclear bombs, aircraft carriers, drones and what have you. It's not some new concept. You can take it as given that regimes you will never want to live under will be putting their best people in using AI for military purposes.

It's dangerous to imagine one is sitting on some moral high ground by refusing to work for the military. We all owe our freedom to previous generations of scientists and engineers who worked for the military, sometimes inspite of strong ethical concerns.

The graph you are linking to ends in 1998, when Bill Clinton was in power, the US was the sole superpower, the EU was increasing in size, China's GDP was a tenth of what it is now, and i think Boris Yeltsin was running Russia, and most people had not heard of Osama bin laden. I think it is safe to say the world is different now.

Look at the backsliding in Turkey, Phillipines, Burma, Venezuela, Egypt. Look at the regimes which have come to power in Hungary, Poland. And of course Russia itself. China, which at one time seemed to be heading for more moderation has taken a sharply authoritarian turn. I would argue just in the last five years things have turned sharply for the worse.

Do we really want to test a theory that we can have a repeat of the 1930s when the democracies fell far behind an autocratic regime in the arms race, and again the autocratic regimes will not win? Democracies are in retreat around the world, and I only hope we wake up before it is too late.

The military-industrial complex may have become far bigger and perhaps in some ways a burden, but the world is a dangerous place and becoming more so. In my view, there is a perfectly valid argument that working with the military is the right thing to do.

At the same time if individuals have pacifist leanings and do not want to work for the military, i would hope corporations respect that.

Long back when this started, OpenDNS was the largest public DNS provider. And I know at one point they started "hijacking" google requests - you can read this here : http://www.dslreports.com/forum/r21284386-Google-Com-Hijacke.... This was in 2008. They used to also hijack DNS requests to show their own ads. And i am pretty sure many ISPs always played around with DNS in this way.

In 2009, Google launched its own Public DNS - the reasons given were to make the web faster and safer, and i think there is some sense in which they have delivered. But I think another reason was defensive - provide an alternative to OpenDNS and to crappy ISP DNSs.

In 2014 - OpenDNS stopped using DNS to show ads : https://umbrella.cisco.com/blog/2014/05/29/no-more-ads/.

It allows them to develop self driving technology without the expense of dedicated cars and paid drivers, and without any liability (because drivers take the risk).

Might be worth reading this article :

https://media.blackhat.com/us-13/US-13-Martin-Buying-Into-Th...

It is really pointless using the count of CVEs as a measure of how vulnerable a product is.

Please see : https://research.swtch.com/vgo-repro - reproducible, verifiable builds. The same thing is possible in vgo, with planned support for hashes.

I believe that vgo get -u works for both direct and indirect dependencies.

See the vgo tour : https://research.swtch.com/vgo-tour

The indirect dependency (rsc.io/sampler - which is a dependency of rsc.io/quote), is also upgraded to the latest version v1.99.99 when vgo get -u is done.