mik1998

Reads very nicely but I can't imagine someone checking the dictionary for every verb or noun they use in order to determine whether it is approved.

Because disk manufacturers can produce less drives per TB if they simply change the measurement. Not to mention they have a whole army of useful idiots who are willing to fight for their profits.

Telling people they need to find you more attractive or else they're racist and evil is a comically idiotic thing to advocate for.

7z is an archive format, it just uses LZMA2 compression I believe.

I don't know about holidays but Israel uses the same daylight savings as the mentioned Eastern Europe countries, ie switching from UTC+0200 to +0300 at around the same time - only difference is that they start daylight savings on Friday instead of Sunday the same week.

I don't think he was from Eastern Europe, but if you want to look at UTC+0200/+0300, in Europe this only includes Finland, Baltics, Ukraine, Romania, Moldavia, and Greece. But notably if you look a bit down it also includes a good chunk of the Middle East, including Israel.

> that the hacker's activity is mostly in Southeast Asia working hours (read GMT+8)

Opposite is true. His activity is focused around midnight in UTC+0800, with significant activity at 2AM in that timezone.

It's a nice analysis but he misses the fact that the Eastern Europe timezone doesn't match office hours, in particular it'd mean he worked around evenings primarily (see this graph https://files.catbox.moe/4itspl.png)

I had noticed UTC+0300 commits in the repository under his name but I believed they might have been simply committed by the main Finnish maintainer who is in the UTC+0300 timezone.

> But I would like to see analysis of timestamp of GitHub events (like PRs and comments timestamps) which are harder to fake.

I doubt the git commit timestamps are faked, since actually faking them is somewhat difficult to do consistently (you would time travel frequently). I don't think there is some kind of github API for this, however from what I've seen they seem to match up with the same work timespan you see in the commit timestamps.

I used the git repo at https://git.phial.org/d6/xz-analysis-mirror/commits/branch/m... and simply did

  git log --format=%cd --author="Jia Tan" 

and then processed it a bit with gnuplot. Should not be difficult to reproduce this graph, but I am not too much of a gnuplot wizard so I first preprocessed this into some different files in a REPL. Don't have the full code of what I did but it should not be difficult to reproduce, just parse the dates and look at the hours.

I did my own research. If you look at the git repository commit log and some mailing list messages, you will see that the author ("Jia Tan", fake name) speaks impeccable English (already lessens the chance of being a Chinese operative), however he commits in the +0800 time zone (Beijing). He works during Chinese holidays and doesn't work during Western holidays.

However, the times don't make sense: It looks like he works mostly at 2am: https://files.catbox.moe/6mdtez.png (hours in the +0800 timezone). I understand this to be indicative of using a different timezone on the computer than where he actually worked, possibly knowing that git commits include the timezone.

If you shift the timezone to US East Coast -0400, it suddenly looks like a very comfortable full-time job, including a fall in commit rate right where the lunch break should be: https://files.catbox.moe/dtvjzr.png

To me, considering that this appears to be a nation-state tier attack, heavily indicates that it was the Americans. Obviously not conclusive proof, but I think it is useful evidence.

Personally, I use lzip ever since I read https://www.nongnu.org/lzip/xz_inadequate.html Seems like the complexity of XZ has backfired severely, as expected.

"could", theoretically. In practice, there has never been an observed exploitation of the supposed vulnerability.

mitigations=off

I'm not sure I would come to this conclusion. R has some adoption, but it's also really not used as a generic programming language, which most Lisp dialects are.

People who produce IP either by work for hire or by selling their copyright rights.

I've had people "steal" IP by sending me a PDF of a paper they wrote whose copyright resides with their university. Not the typical image of a thief.

Absurd, presumably anecdotal claim.

My own anecdotal experience is that it has no relation and I know many people who have violated the copyright on the IP they themselves produced.

I often use a Thinkpad X220 (which still works for a lot of my usage and I'm not too concerned about it being stolen or damaged) and the JS web is terrible to use on it. Mostly resulted in my preference of using native software (non-electron), which generally works perfectly fine and about as well as on my "more modern" computer.

I mean... most European countries (modulo some exceptions like Germany) existed (non-continuously) for at least a thousand years.

It's an infinity% more than what I paid for university in Poland

Frankly, I never understood the anti-homework crowd. In university, I enjoyed homework - I am too stupid to do exercises quickly/during class, and homework lets me think about things for a lot longer. Doing homework has correlated very heavily for me with actually remembering things later on.

In high school and before homework took minimal amounts of my time. I am for assigning more homework to teenagers generally, since they do not do worthwhile things in their free time, as long as the homework is reasonably complex.

Guix has this feature - guix pack https://guix.gnu.org/manual/en/html_node/Invoking-guix-pack....

In Europe unlike in America the government makes regulations that benefit the general public. That's it.

The EU should fine as much as it is necessary to financially impact the corporation for violating the laws.

Aren't they literally Ossi?

It's nice for general numerical methods but somewhat annoying for implementing integer-arithmetic algorithms.

PRL was not "Soviet occupation", unless you're referring to parts of Poland during 1939-45

It looks like regression to the mean.

Static typing is useful if you are building some program for other people to run that you want to be reasonably robust. However, it does not work well with interactive programming. If I had to declare all the types every time I do something in Mathematica, I'd be severely annoyed for no benefit.

Common Lisp seems to be the best of both worlds, with dynamic strong typing in general and advanced compiler type inference that can correctly point out mismatched types in most cases in SBCL, leading to much easier time programming interactively.

A statistical model cannot understand anything since it does not think in the first place.

Are you claiming that Russia or China are paying 4chan moderators? That seems unlikely for an American website owned by a Japanese millionaire.

Apparently one of {Croatia, Serbia}. I think we need to consult the axiom of choice on this one.