ekns

Plugging in a past question I had related to this [0]: Given that ECC RAM is not abundant, why don't operating systems have self-checks for RAM? Instead we only tend to have near-post mortem checks with memtest86 and the like.

[0] https://news.ycombinator.com/item?id=24642062

I wonder how they're cooling this. That's 10 electric kettles worth of heat output :)

Currently I'm using namesilo as a registrar, and my own domain for email, albeit with G Suite.

I hope that with G Suite you won't get kicked out arbitrarily, but I think I could still recover from that. Hopefully the registrar will never need email confirmation to log in (assuming my DNS setup is there).

Still, I'm worried that could get me kicked out of my registrar illegitimately. Is there anything that can be done to mitigate this aside from using something super expensive like MarkMonitor?

Earlier my account at namesilo got disabled silently when they decided by themselves that my corporate credit card was somehow fraudulently used by me. I got it resolved eventually, but I fear something similar could eventually lock me out completely.

FWIW, I just found out there's a tool called memtester [0] that could fit in here. But then, I wonder why operating systems don't use it. IMO they should integrate these things so it's done systematically for most users unless explicitly disabled.

https://shipilev.net/jvm/test-your-memory/

One of the other submissions relating to this became flagged for some reason: https://news.ycombinator.com/item?id=23526311

I don't often comment on things, but this seems like a slippery slope and a process that I'd like to counteract in some way, so input/comments are welcome. Original comment copied below:

I'd prefer it if things didn't change, including the master/slave terminology (the 'master' branch doesn't even have a 'slave' in git...)

Have the words themselves become bad to use? It seems that this is a USA centric view anyway with its history (and even then it seems ridiculous--why can't these things be compartmentalised as technical terms?)

Slavery certainly still exists in the world, and has existed in most places at some point. Should we rename 'slavery' to something else too while we're at it? Or forget it exists and existed?

Certainly certain terms become antiquated over time, like the vocabulary used in Emacs (yank/buffer/etc vs cut/window/tab/...), but master/slave doesn't seem to be there yet.

I wonder if there's some way to counteract this newspeak stuff? I think it's harmful that these things happen for such frivolous reasons. Does one just have to be as loud as the minority that drives these things, to counteract it?

I'd prefer it if things didn't change, including master/slave terminology (the 'master' branch doesn't even have a 'slave' in git...)

Have the words themselves become bad to use? It seems that this is a USA centric view anyway with its history (and even then it seems ridiculous--why can't these things be compartmentalised as technical terms?)

Slavery certainly still exists in the world, and has existed in most places at some point. Should we rename 'slavery' to something else too while we're at it? Or forget it exists and existed?

Certainly certain terms become antiquated over time, like the vocabulary used in Emacs (yank/buffer/etc vs cut/window/tab/...), but master/slave doesn't seem to be there yet.

I wonder if there's some way to counteract this newspeak stuff? I think it's harmful that these things happen for such frivolous reasons. Does one just have to be as loud as the minority that drives these things, to counteract it?

Switched to Firefox recently on my desktop. I had the strangest problem with Chrome: Somehow it just kept forking more and more processes until the rlimit got filled (100k+ threads/processes).

Too much trouble to debug it so I opted to switch instead.

Hm I was doing something similar, but in the end there was just too much stuff that's broken, all the time: https://twitter.com/everythingisbr2

In the same vein, I've been thinking of having some universal bug/annoyance reporting tool where you'd tag each annoyance with the software/hardware/website and optionally its version.

Pandoc only handles a subset of org-mode markup.

IIRC last time I tried this (long ago) there were a number of gaps in things like org-mode tables.

At the limiting case, the (common or uncommon) elisp extensions (e.g. org-babel) will likely never be processed by third-party tools.

Failure 1: I'd planned to do a PhD eventually, but I ended up more or less self-sabotaging my degree grade by being depressed and not doing anything about it (2:2 in the UK system vs. the nominal 2:1 ostensibly required for MSc/PhD applications to most places)

Failure 2: For the longest time, I never managed to make a good routine for myself. To learn, build things, exercise, meditate, etc. Good habits seem to be a pre-requisite for success in virtually any endeavor.

Failure 3: I had an interview scheduled with Google once, right after I had finished my military service in Finland (it's more or less mandatory for men).

I had scheduled it a tiny bit too close for comfort, just after I was supposed to have been out of the garrison. Turns out they called an hour too early (I specified UTC timezone but they marked it down as BST..) so I could only awkwardly tell them I could not talk at the time, and ask if they could call again an hour or a day later.

I never could get another call and settled for a crap job (by some standards). Maybe I could have earned a lot more and learned a lot more (and had a better network of contacts) if things had turned out differently :p

On a more positive note, I started doing freelancing/consulting last year after "no one" seemed to want to hire me at a decent salary and proceed to make OK amounts of money through random gigs.

Still figuring out how to network effectively. I figure the best bet is through gaining visibility online by publishing something useful.

How is the universe infinite by definition?

Wikipedia for one says the following. There is no mention of infinity.

"The Universe is all of space and time[a] and their contents,[10] including planets, stars, galaxies, and all other forms of matter and energy. While the spatial size of the entire Universe is still unknown,[3] it is possible to measure the observable universe."

All or most Finnish banks use an ~8 digit account ID and a 4-6 digit one-time password from a slip of paper that has 80 to 300 of them. Some have a separate PIN/password too.

For confirming large transactions (>5000 or 10000 eur) there's a separate phone call or SMA verification.

Nowadays there's apps for 2FA instead of always requiring the use of a one-time password. My corporate bank account still uses the paper backed one-time passwords though.

I've never really like the word "anti-pattern". Anti-patterns are still patterns of sorts instead of pure entropy.

How about just "common mistakes" or "discouraged/deprecated/error-prone/incorrect patterns" or something?

> The ultimate solution to IPv4 exhaustion is, of course, the complete transition of the Internet to IPv6, however, this will take time and until then there will be (by definition) networks and sites which only support IPv4.

Off-topic: I've always wondered why English speakers (or perhaps just Americans?) use "by definition" arguments so much (also just saying that things are by definition so and so).

People never seem to use "by definition" arguments and such in Finnish for example.

Definitions of mathematical objects aside, things in concept space are not eternal and can shift around. Is the usage a cultural thing? A quirk of language? Just some random trivial thing that just is and doesn't have any particular reason to it? :p

It's not even slower (and it's safe beyond the argv limit) if you use "+" instead of the semicolon.

  find -type f -exec sed -i s/foo/bar/ {} +

This only invokes the command twice for 100k files (assuming we can pass in 64k arguments to sed)

Location: Helsinki, Finland

Remote: preferred

Willing to relocate: yes

Technologies: Python, Flask, modern SQL, Postgres, MariaDB, Elixir, Phoenix, Jupyter/Pandas/etc, ES6+, Haskell, Java, AWS, Google Cloud, etc.

Résumé/CV: kunnas.com/cv

Email: hn at kunnas.com

Data science, full stack development, system design. I am interested in end to end design space exploration/optimization and software correctness from a pragmatic point of view, maximizing ROI, minimizing technical debt--QuickCheck beats ad hoc cases, correct by construction beats tests, design optimization may make the code (and its testing) superfluous entirely.

There's something similar in Finland from a company called Visma: https://allekirjoitus.visma.fi/

It's tied to the de facto standard bank/other authentication system called TUPAS[1] which is used to prove your identity to government services, banks, etc. etc.

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

Unfortunately org-mode is not backwards-compatible and there have been many changes which break things unnecessarily.

For example, I once attempted to run this org-mode "notebook", ironically titled "Reproducible Research with Emacs Org-mode", and found I had to make significant cosmetic changes to get it to build: https://github.com/eliask/orgmode-iKNOW2012

Maybe things have improved since, but backwards compatibility is important for these kinds of formats.

I recall seeing this Google Tech Talk by Steve Hsu on how they are in the process of finding out basically all the additive genes that contribute to intelligence in China at the BGI research institute: https://youtu.be/62jZENi1ed8

They would take roughly a thousand people who are >=3 standard deviations smarter than average and compare their full genomes against people who are of average intelligence. This way they expect to find out all the "easy" genetic determinants of intelligence.

  Location: Helsinki, Finland (atm, I travel a lot)
  Remote: Preferred
  Willing to relocate: Probably not
  Technologies: Python, Haskell, Elixir, Postgres, Tensorflow, keras, etc.
  Résumé/CV: kunnas.com/cv
  Email: hn@kunnas.com

AI-flavoured full stack developer :)

Even weirder still, in my experience, is that Google Maps randomly displays the names as Finnish sometimes, and randomly displays them in Swedish other times.

I've been happy with gnome-terminal :)

I figured out how to use the Falcor API of Netflix to fetch a full list of genres.

Submitted to HN separately, but here you go :) https://eliask.github.io/netflix-genres/genres.html

The repository contains the scripts to fetch them also.

Random examples from recent experience:

Some exist in Finland (at least in Helsinki).

Also saw one recently in a train station in Belgium, and IIRC a train station in the Netherlands.

Also in the North of Argentina near some bus station, where you would pay directly to someone apparently responsible for cleaning/maintaining them.

Improved version using jq. This properly URL encodes the query parameter and uses a much simpler sed command. Additionally the API returns UTF-8 encoded data when the user agent is specified. Requires curl >= 7.18.0.

    function c() {
        url='https://www.google.com/complete/search?client=hp&hl=en&xhr=t'
        # NB: user-agent must be specified to get back UTF-8 data!
        curl -H 'user-agent: Mozilla/5.0' -sSG --data-urlencode "q=$*" "$url" |
            jq -r .[1][][0] |
            sed 's,</\?b>,,g'
    }

Example:

    $ c ':)' ':('
    ) ( meaning
    ) ( emoticon
    ) ( ͡° ͜ʖ ͡°)
    ) ( emoticon meaning

I had surgery with my first salary more or less when I was 21. It took a year or two for the halo effects to disappear.

My night vision feels like it's like "4x" worse than it it was but it is really hard to tell. Seeing in the dark is difficult in the first place and I don't have any hard data.

Funny how it says there's 479830 pages worth of results and even has a link to that "last page".

I checked and in practice it "only" returns the first 400 pages of results.

I recognize the syntax in "date_trunc('day', created_at)::date as day,", namely the cast "::date", as PostgreSQL syntax. Could be something else as well but I've only seen it with PostgreSQL.

Even if you could buy one, one of them costs 12k euros last I checked (a few years ago)

Not sure if this has changed with more recent models, but I've had rather bad experiences with UX31.

Although the model number was the same, most of the European variants had a different, extraordinarily bad SSD, which slowed down to <1 MB/s if the file system was something like >70% full.

Other indicators of bad quality:

- keycap texts disappeared for the most used keyboard buttons after just a month's use

- crashed once every three days on average (hardware issues?)

- The fan makes a buzzing sound as it scrapes the bottom of the laptop casing

- The screen died after about a year of use (booting up makes the screen gradually turn completely white within 5 or 10 seconds)