edejong

In which way exactly? Having worked with many, many different package managers, I've come to appreciate Maven's way of doing things. Even though it has many drawbacks:

- XML structure is verbose

- Plugin architecture makes standard behaviour opaque

- Default HTTP with no package signatures is recipe for security problems

- "Stages" seem arbitrary and it's not always clear which stages are performed.

- Community seems to move to newer systems such as Gradle (and SBT for Scala).

But, there are also strengths. The configuration format can be extended easily using the plugin architecture. This provides a huge opportunity to extend the functionality.

It is battle hardened over time, which means it is very stable and it has a huge plugin architecture.

I hope you’re /s

The established car manufacturers are too highly optimized for their specific suppliers. Over many decades an internal culture appears within these companies that immunizes against heavy risks. This is done by (unconsciously) selecting for certain personality traits in upper management and by forming certain specific ties to the first and secondary surroundings of the company.

In short: companies petrify around local optima given time.

Yes. Fun emerges due to sharing of similar goals. People have fun because they play the same game, enjoy the same movie or simply have the same freedoms. I agree that it helps direct and motivate us. (Thanks for that insight!)

I guess I just cannot share in the purpose of the team making this car go +300 mph. We already have the Thrust SSC (1228 km/h / 763 mph). So what does this prove? That if you want to take corners and need to see what happens behind you, you won't get to go that fast? Why not productionize the Thrust SSC then?

Max speed (to me) is just a senseless metric. It's a bit like clock speed of a CPU. Interesting metric, but only in relationship to other factors.

No, because a dragster has a specific function. It is a perfect counter-example to the senseless: 'look how fast our super-car can go.' There are many more values involved in dragster racing: knowledge, innovation, joy, creativity, competition, determinism, meaning, enthusiasm and playfulness. And it is all related to a singular goal: getting over the finish line faster than the competition.

But with such a super car, idk, I'm just left with an empty feeling: why? What is the point?

Yes, but there is an interesting difference between the two w.r.t. status.

Status is generally derived from the signaling of certain values. Wealth is one such value, but in general we assign status to a combination of values. Aesthetic + Wealth = Art. Strength + endurance = Athletics. Leadership + Vision + Knowledge + Intelligence = Innovation.

For me, at least, these super expensive cars have lost all signaling value, except 'wealth'. Because of that, I would almost call it tasteless, senseless and idiotic.

idk... I've always assumed that the pointlessness is in the playfulness: extending our boundaries, setting new rules. We value the daring, the boundless, the discoverers and the challengers. And I understand we connect a certain status to that and I also understand that certain people just want to be able to buy the status without putting in the time, knowledge and risk. But for me at least, this is so remote from and unconnected to the original challenge and thrill, that it becomes idiotic: pointless without even the playfulness.

This got me thinking. Once I thought it was fun, but the more I look at it, the more I feel it is utterly pointless.

A racing car has a function: to get around a track faster than other cars (over a period of laps). To do this, it needs to be versatile: turning, braking, acceleration, fuel consumption, range of vision and suspension (among others) play a role. That's the true engineering challenge. It's why it has wheels, not wings. But to simply go fast, seems a bit pointless to me.

A supermarket is not a public space. All I am saying is that all information gathered in the public space should be shared publicly. The idea being: anyone could have been there and recorded that information (since it is public). This will prevent those with deep pockets to create a panopticon of our public environment.

> At least right now, that's not true. It's harder to automatically scrape Facebook than a random decentralised system - mainly because decentralised systems need to have the data sharing built in to work. Sharing the data in a structured way is the central idea behind them.

I disagree. It is harder for Joe Average, but for state actors MUCH easier to scrape one system, instead of thousands. Also, it is possible to design decentralized systems in such a way that scraping becomes almost impossible.

> You're not going to be able to block this access from gov only. If known networks become blocked, new gateways can be obtained for less money than is worth mentioning.

You can block based on more than IP address. Decentralized systems can use networks of trust to curtail unwanted information gathering.

> Again, if you're interested in anything resembling a social network, you'll need to share your data with a lot of people. Unless you want to keep it all private - you can do the same thing on FB currently, and border control will not have access to it.

Of course border control has access to it, regardless of your privacy settings. Do you really believe FB has full control over your data?

Ah, I see now. Back to your original comment:

"To prevent cities being turned into panopticons, we should make sharing data with people trying to build them mandatory"

The reason why panopticons work is because the prisoners don't know what the guards can see. And due to this asymmetric power relation, it is very hard to revolt. There are two solutions: either prevent the information gathering, or make the information publicly available. The former solution is not possible if you don't know what information is being gathered, so we are left with the second solution (to start with).

Quote from?

Regulation proposal: all video data and derivatives (face recognition) of public areas must be publicly available. Sort of GPL-like. The resulting effects will force people to rethink and reevaluate current policies.

Your freedom is primarily based on the amount of power you decide to give to others. Due to social media, a tremendous asymmetry of power has been created in the hands of a few. This power is made visible in areas that are slightly outside of normal jurisdiction, such as at US border control.

Our normal tendency is to want to regulate, but this can only be done to a limited extend. Too much regulation can stifle innovation or place too much power in the hands of the regulators.

I'm not arguing against regulation. I am arguing that we should understand that our social media use (among others) comes at a very high cost to our freedoms. Proportional use, (proportional to our ability to regulate or our willingness to provide asymmetric information power) is warranted.

Yes, it certainly does, for a variety of reasons.

First of all, it is much easier and cost-effective to gain full knowledge of the graph data in a centralized system. Instead, by working on a distributed system, we could change the protocol to increase the effort needed to mine data.

Second of all, we gain the opportunity to know who or what has requested our data. We can decide to block these requests if they are used against us.

Lastly, we can distribute the available power of information over multiple nations. When policies change, data (and information) can flow to places where these policies have no effect.

With great power lies great responsibility. People give great power by using centralized social media from a single nation without fully understanding the implications of it.

Maybe now people will stop using social media. Ah, wait, no. Everything in the name of convenience.

IMHO, publicly gathered information should automatically be made publicly available. That should give people something to think about.

That. They could’ve just supplied a default .profile / .bashrc. It’s these kinds of bells and whistles that causes most security bugs due to additional attack surface. Is it the root user fetching the news with wget/curl from a remote source?

But when is a tool specialized? Would a dedicated XML / JSON / YAML / ASN.1+DER / Avro / Protobuf / Parquet editor be specialized? And what if that hierarchical standard would be the de facto industry standard? Is a binary file editor specialized? What about a structured assembler?

Personally, I think most editors are rather specialized. They deduce the character sets, often add syntax highlighting and provides paging for very large files. High level strongly typed languages, such as modern C#, Java, Scala are designed to be used in an IDE. You could view and edit it, but it provides a difficult situation (not unlike editing XML by hand).

"Human readable" is very subjective. It depends on the person, the task at hand, the intended recipients, etc. etc.

Perhaps you should stand on that soapbox a little more often. People are way to enamored with their little 1970's ed and ex derivatives.

Every veteran in the field knows that data and data structure are the primary enablers for almost any solution. Nevertheless, we have regressed in the last decades w.r.t. that. Had we had better structured (AST-driven) editors and context dependent representation, XML (or sth similar) might have taken off.

Personally, I blame the exponential increase in inflow of new developers. Nothing but reinventing the wheel without knowing history.

No, voting is passive. Be politically active: talk to your neighbors, at your local sports club, at work and at local events on climate change. Contact your local legislators. Give options to local polluters by researching options for subsidies. Write to your congressmen and ask for the most pressing local actions.

Democracy does not start with a vote. That’s just a step in a large process.

Would be nice to contrast this with the energy necessary to produce 1 kilogram of various foodstuffs.

Another problematic bias from the article: "Respondents stating that they used an e-bike were categorized as e-bikers, independent of whether they also used a conventional bicycle."

So, since belonging to a group of bicycle use (leisure, sports, commute) is additive in the amount of exercise one receives, the e-bike group is automatically larger than the non-e-bike group.

How does this pass peer review?

Article: "Users who reported any use of non-electric bicycle (including city bike, mountain bike or bike-sharing) but not e-bike were categorized as cyclists, those who did not report any bicycle use, as non-cyclists."

Aha! So, we're comparing mountain bikers and bike-sharers with e-bikes (which belong in the category of electrified city bikes). That alone should raise some informed eyebrows.

And we shouldn’t believe for a second this isn’t already happening with multiple three letter agencies.

No, accuracy is but one possible goal. "Machine learning practitioners take pride in models that optimize the value of that model."

And since we cannot define exactly what this value is (accuracy, RMSE, bias or a specific confusion matrix), we have to make a more abstract definition.

We are now left with a question. Is this abstract definition quantifiable? If so, we can still exclude the statistician w.r.t. understandably. However, if we allow qualitative value as well, the traditional statistician is back at the table.

Now, a further refinement needs to be made. Since fairness, accountability and transparency in Data Science are in the limelight, we can make a point that our ability to understand is a key metric for Machine Learning models as well. It is interesting to see the tendency to associate words or phrases to key neurons in embeddings. For example, in embeddings of faces, we can associate gender, ethnicitity, fatness and gaze with some of the embedded neurons.

I’m just waiting for the day when Tinder is down in popularity, security fixes are a bit more lackadaisical and a zero-day exposes a decade of personal preferences of a large share of the population, not unlike a nuclear waste leak. Imagine the awkwardness when a coworker finds out you swiped them left (or right).

And no, I am not going to end this with a paternalistic or moralistic statement.

Currently running 4 JVM based solutions on 2 RPi's in our home. One is a fairly heavy one (OpenHAB 2), consuming 700 MB of memory. The others are simple importers (written in Scala), using around 50 MB each. It runs fast, is extremely stable (> 200 days uptime), very fast response times (< 100 ms), has a built-in dependency manager and is type-safe.

With this setup, I have aggregated more than 4 years of in-home data, ranging from radio-thermostats, 15 thermometers, continuous electricity usage, around 10 z-wave devices, 15 zigbee sensors, Philips Hue Lighting (8 bulbs, 8 switches), local weather updates, InfluxDB storage and a fairly complicated setup for climate control (including presence detection using OwnTracks / WiFi and operation of roller-shutters and sunscreens).

The question was whether it is hard to determine the space/time complexity of ‘a’ function, which firmly placed the question in Rice’s theorem grounds.