azag0

Yes, but then the whole ecosystem is missing micropayments. I subscribe to some sources, but sometimes I just want to read a single article. Would gladly pay 20 cents, if it was a matter of a click.

Yes, but Paypal nor Walmart don’t handle irreversible transactions. Nor are they platforms on which apps and custom transactions can be implemented. So I don’t think this is a good analogy.

The new problem here is that the centralization is pretty much the core value of these companies. Breaking them up would destroy that value. A way around is to first standardize and open all the protocols on which they internally operate, and then split them. But you could say that the standardization is in a way an indirect nationalization. If email was run by a company, it would easily be amongst the richest companies in the world.

As a python package developer, I'd assume that being able to install with pip in user space or in a virtual environment is a standard technique. Why duplicate the dozens of tutorials on the internet?

Correct.

Rather than meaning “superb conductivity”, superconductivity is a specific electronic state of matter in which the conductivity drops to zero at a certain critical temperature, and magnetic fields are “expelled” from the material. In conventional (low-temperature) superconductors, this is caused by formation of electron pairs (Cooper pairs) that behave as bosons. There is no established mechanism for unconventional (high-temperature) superconductivity.

In any case, the present study does not mean that graphene is a superconductor.

I think much effort went into understanding why AlphaGo makes the decisions it does, and it’s almost a research project in each particular case.

Very trivialized: in some sense, you could say that for light itself, there is no time. In the same sense as there is no space for things that do not move (in space).

In chemical physics, many of the most influential papers (tens of thousands of citations) have been published in The Journal of Chemical Physics, an absolutely non-flashy publication with impact factor ~3 and a myriad of uninteresting papers.

I use rsync [1] to backup my laptop to my server twice an hour. (Hard-linking files that already exist in previous images.) Then another script [2] on the server prunes old backups every night.

[1] https://gist.github.com/azag0/187174485a35a70d9d1d79c8533a6b...

[2] https://gist.github.com/azag0/b975c8a1935db29376e20926404d00...

Bitcoin mining is trivially parallelizable. Most supercomputers are built for problems that are not.

You’re not alone. Took two years (until now) for me.

From what you write and from the diagram in that pdf, it's the issuer bank that creates the device token, but then it's the network that does the reverse lookup. Do both the issuer bank and the network know the mapping between device tokens and cards?

It’s really more about payment cards (debit, credit) in general, not credit per se.

Where exactly does Apple Pay fit in this chain? Do they replace the card issuer or are they a new link between the acquiring bank and the issuer?

Perhaps the reason is to prevent anything from spying on the clipboard. Keyloggers are likely prevented by Secure Keyboard Entry. In contrast, websites cannot spy on clipbaord, so there is not real reason to prevent paste on websites.

I agree. I think the closest word from the nondigital era to a "blog" would a "bulletin".

I’ll say it again. If the natural language of your problem domain is linear algebra and/or matrices (tensors), Fortran is the perfect tool.

People now try to use ML anywhere and everywhere so it's wild west a little. Three examples: [1] uses a standard neural net to represent a many-body wave function, with all the machinery of quantum mechanics on top of that, and reinforcement learning to find the true ground state. [2] uses a handcrafted neural net, which by construction already takes advantage of a lot of prior knowledge, to directly predict molecular energies. [3] uses a simple kernel ridge regression coupled with a sophisticated handcrafted scheme to automatically construct a good basis (set of features) for a given input, to predict molecular energies.

In all these cases, the ML itself is not the target problem, but only a tool, and most effort goes into figuring out where exactly to use ML as a part of a larger problem, and how to encode prior knowledge, either via feature construction or neural net handcrafting.

[1] http://sci-hub.io/10.1126/science.aag2302

[2] http://sci-hub.io/10.1038/ncomms13890

[3] https://arxiv.org/pdf/1707.04146.pdf

This is a huge topic in applying ML in physics and chemistry where we already have a lot of prior detailed knwoledge about the systems we want to describe and it would be silly not to build it into the ML models.

I ended up writing a simple distributed build system (https://github.com/azag0/caf, undocumented) that abstracts this away. (I use it at a couple of clusters with different batch systems). You define tasks as directories with some input files and a command, which get hashed and define a unique identity. The build system then helps with distributing these tasks to other machines, execution, and collection of processed tasks.

Ultimately though, I rely on some common environment (such as a particular binary existing in PATH) that lives outside the build system and would need to be recreated by whoever who would like to reproduce the calculations. I never looked into abstracting that away with something like Docker.

(I plan to document the build system once it's at least roughly stable and I finish my Phd.)

Was the comment redacted?

I tried it and was mostly happy with it. But has it gone through anywhere near the third-party scrunity that 1p has?

I am a scientist who sometimes publishes data sets with academic papers and this looks super useful both as a tool and as a potential publishing best practice. Currently doing away with HDF5 and figshare. One necessary future for academia would be to be able to assign a DOI to a given version of data. Is it feasible for quilt to have such a feature?

As does any regulation. That does not mean the regulation is bad. I wouldn't regulate this issue with specific rules, just put a cost on waste. Each company would need to deal with the product waste in an environmetally safe way. If they can essentially recycle the replaced products into new ones, then I have no problem with that. But if it ends up as waste, then there should be a price on that. I wonder how much would this motivate companies to make repairable products.

The teenage memories of hours spent tweaking this lovely program... Was the software I missed most when I went the Apple path.

The yearly budget of arXiv is ~$1M.

I don't see a reason why science couldn't flourish on a non-anonymous message board such as reddit or twitter in place of peer review. This is all matter of momentum. And of fear of open criticism.

I have much doubt that criminal penalties would change anything. People would just put much more energy in avoiding them. I think that state a priori regulation is a more effective tool than state a posteriori punishment.

Some parts of physics might stretch the ELI5 to its limits. I like to explain science to my non-science friends, but I'm always hitting the issue that parts of the specialized knowledge become intuitive, and then it's difficult to go back. Say, my understanding of how electrons behave in matter is a category on its own (let's call it quantum), and it's difficult to translate to common terms.

You could replace "plain text" in the article with "non-binary", and it would probably make more scence. Markdown is also not plain text in the most strict sense. Or HTML is plain text in a vague sense. In the end, what really matters is how easy it is to build parsers and tools for a format. Being non-binary is a huge plus. I think that was the point of the article, and I agree with that.

This is the old security via obscurity. At least though, one should know that one publishes the full source.