michaf

If you are interested in polyglot files like this, have a look at the work of [0] (who was probably directly involved in this particular release).

[0] https://twitter.com/angealbertini

Actually telcos in those countries with many foreign visitors will still make money. While consumers enjoy 'roam like at home' without additional costs, there are still wholesale roaming charges in place for the home carrier [0]. A quick back-of-the-envelope calculation shows that these wholesale roaming caps are comparable to typical consumer prices, meaning they should result in actual profits (at least compared to German prices, which are generally not the cheapest).

[0] https://en.wikipedia.org/wiki/European_Union_roaming_regulat...

There is a newer version 2 from 2014: https://arxiv.org/html/1202.4212v2

I do not see that option anywhere.

Sorry to hear that. For me the only side effect was some drowsiness, but I heard that it can come with more severe side effects for others. Best of luck with your further pain management!

Probably you know about the opioid alternatives, but did you try anticonvulsants [1] for your nerve pain? I had nerve damage after a surgical procedure, and opioids did not really help. A doctor suggested to try Gabapentin, which greatly reduced pain down to tolerable levels. While anticonvulsants probably have their own risks, they seem much less addictive than opioids, and are probably easier to get on prescription.

[1] https://en.wikipedia.org/wiki/Anticonvulsant

I just constructed a little POC for bittorrent: https://github.com/michaf/shattered-torrent-poc

Installer.py and Installer-evil.py are both valid seed data for Installer.torrent ...

I wrote a small proof of concept here: https://github.com/michaf/shattered-torrent-poc

Both Installer.py and Installer-evil.py result in the same piece hashes, with piece size chosen as 512. The only difference between the files is the shattered pdf fragment, encapsulated in a variable. This fragment starts exactly at the piece boundary, resulting in identical hashes.

nim is faster at calling c routines than c? That seems interesting... But probably the measurement errors are larger than 1%?

You are correct. It seems they really should double-check their implementation for producing meaningful results before running bechmarks on it...

This seems to have been fixed 3 hours ago: https://github.com/marblestation/benchmark-leapfrog/commit/c...

gcc seems to do something similar for -Ofast. Adding something like

  printf("x[0][0]=%f\n", x[0][0]);

at the end of main fixes this.

Benchmarking is hard.

Do you participate in the current Data Science Bowl regarding CT lung cancer detection [0]? The prize pool of $1,000,000 seems quite attractive, especially if you recently developed new state-of-the-art CT lung cancer detection ML models. The only somewhat strange aspect of this competition (at least to me) is to not include locality annotations. They only provide labels of cancer/no cancer per patient...

[0] https://www.kaggle.com/c/data-science-bowl-2017

I agree. I see it as an attempt to somewhat deflate the emerging bubble. Most AI researchers are aware of the effects of the previous 'AI Winters' and probably want to smoothen the hype cycle in order to secure continuous funding in the future.

Super fun game! I only figured out the importance of using WASD (i.e. chaining) in my 3rd game. Maybe you can make it more clear in the tutorial why it's a good idea to use them.

> What we can count on is that MacBook Pro prices won’t get any higher for many years

This is pure speculation stated as if it were fact. It makes me question the validity of the rest of the article.

Interesting read. Researchers in the HPC community have developed a number of performance models to predict real-world performance in more detail than possibe through simple Big-Oh of number of operations, e.g. while OP concentrates on latency, the Roofline model ( https://en.wikipedia.org/wiki/Roofline_model ) mainly considers limited memory bandwidth.

I had a similar experience. Increasing pain levels gradually drove me away from from pursuing programming projects outside of work. First I tried the Kinesis Freestyle2, and while the pain got better, my hands could not adjust to the additional degree of freedom from having two keyboard pieces. Got the Advantage, and now I can code again for hours without any wrist pain. Highly recommended.

This reminds me of the 2010 Digg redesign, which resulted in a massive user exodus to a much less polished looking reddit...

That makes $60.5 per user. Compared to other social networks this seems to be within a reasonable range.

Nicely done!

Minor nitpick: your wait-for-dynamic-content spinner (https://remotebase.io/images/logo.png) exhibits eccentric rotation, which might make some people a bit seasick :-)

I used latexmk to streamline the generation of my PhD thesis. I must have run your script literally hundreds of times. So, thank you!

Location: Germany (Frankfurt area)

Remote: yes

Willing to relocate: yes

Technologies: Python, C++, scikit-learn, TensorFlow, VTune, MPI, OpenMP, git, bash, HPC

Résumé/CV: https://www.linkedin.com/in/mfeldmann

Email: job@feldmann.cc

Jupyter Notebooks are great, I use them a lot for scientific computing and data analysis.

The critical missing features in my opinion are 'undo' and proper integration with Git (or another VCS).

Location: Germany, Frankfurt area

Remote: Maybe

Willing to relocate: Yes

Technologies: Python, C++, machine learning (sklearn, caffe, some torch7), OpenMP, MPI, math+physics

Résumé/CV: please request per mail

Email: job@antinull.de

I am a consultant for scientific high performance computing, with a keen interest in ML and a PhD in physics; looking for algorithm- and/or data-heavy development positions.

Location: Franfurt area, Germany

Remote: Yes

Willing to relocate: yes (EU, US)

Technologies: Python, C++, compute clusters, OpenMP, MPI, git, vim, caffe, scikit-learn, VTune, Matlab

Résumé/CV: please email me

Email: michaf -at- antinull.de

I am currently working in massively parallel scientific computing/High Performance Computing/Optimization. I have experience in writing physics simulations, data analysis and experiment design in nanoscale physics from my PhD. Looking for positions in ML/data science/challenging dev projects.

How does Keras compare to Lasagne [0], which is also Python/Theano based, and which was used with some impressive results [1]?

  [0] https://github.com/benanne/Lasagne
  [1] http://benanne.github.io/2015/03/17/plankton.html

The next step could be to crowdsource the design of challenges or puzzles. Like a 'masters tier', where you get special points for designing particularly clever puzzles. This could help remedy the constant demand for new puzzles due to writeup contamination.

BTW, I completed microcorruption, foobar and two stripe CTFs. It was lots of fun and I am really looking forward to Starfighter!

So you could MitM between superfish and the internet, insert

<meta name="superfish" content="nofish">

and everything is all right ;-)

Location: Frankfurt area, Germany

Remote: Yes

Willing to relocate: Yes

Technologies: Python, C/C++, R, MPI, OpenMP, CUDA, scikit-learn, caffe

CV: Please request by mail

Email: job -at- antinull.de

I finished a PhD in Appiled Physics in 2013 and currently work with scientific codes in high performance computing (think 1000+ cpu-core physics simulations). I am looking for a job in Machine Learning/Deep NNs/Computer Vision or similar.