ianbooker

That was a theory I had: That Unicorn leaks and that Unicorn Killer is not killing any of the workers, since it does so all 160 requests and there are no request happening at the moment.

Then I saw the new installation uses Puma ;)

64 GB on a 16 core machine that has nothing else to do.

If any, I think this is maybe too much, every worker just spawns more and more workers. I restricted Postgres from taking all RAM, Sidekiq and Puma to low numbers. But there are more..

I follow the instructions straight from Gitlab.

My new hobby is installing vanilla Gitlab CE on a vanilla server and waiting until it crashes the whole thing. Takes something between 8 to 36 hours.

The server is completely gone, no ping, no ssh. It takes frantic remote power button pushes to even turn it off.

Afterwards I reconfigure various config lines and repeat.

Offer paid support. Most managers in businesses love to outsource accountability for cash.

The trick is to not also take in their work (i.e. for integration).

I am curious: When you 'zoom' in on a single word and read it letter by letter: Is that still happening?

We had a quite strong formal / theoretical community in our cs department, and besides really enjoying automaton theory in my later studies, it kind of formed an understanding of the practical side of researching algorithmic questions like "how to fit multiple circles into one bigger circle to maximize their combined diameter", as a reservoir of solutions you could later abstract your less abstract cs problems to.

The article is from 1996. And the experts then describe the then software industry as "hunter and gatherers".

I am not sure where we are now, was OOP the dark middle age or the industrial revolution? Or will that AI be?

Would be really interested in your opinions!

You can compile such a list for any operating system in 2020.

Recently I evaluated three randomly selected HDMI capture devices, Ubuntu supported each of them out of the box, while Windows and MacOs struggled. With the printer in our office its the other way around. The printer at home is old, so its just Linux supported. And so on..

If I remember correctly, in the early Days of OpenTTD you basically were lost without the files from the original. It took quite a while for the artwork to catch up and finally surpass the original.

I second that. Left the page as soon as I saw it.

I love how this is a total geekery for fun project and one of the best quasi job application in 2020.

The day has come. Last year or so ;)

Macbook, Ipad Pro, Cheap phone (Moto G / Pixel 3A) everything on my desk shares just one power brick.

Or like https://watchnebula.com/ be owned by creators in the first place and try to finance by subscriptions without ads. But this obviously will not scale like Youtube..

Thanks for sharing.

I do exactly the same, and I always find it hard to justify base-R in front of my students, after they learn about the Tidyverse ;)

As so often, these big survey projects lag considerably. At some point in time, these dimensions were most interesting, and it must have been very satisfying to see them explain such a big amount of variance in such a low dimensional space.

I recommend the time series video: https://www.youtube.com/watch?v=ABWYOcru7js

Indeed it seems less a good explanation nowadays, but this is another take away, I guess.

These are well tested social science concepts. Its just hard to communicate in a two word label..

As far as I understand, this is a combined / linked survey and antibody study. So evidence for this must be originating from the questionnaire.

How could you proxy for viral load? And more specifically: How reliable can this measure be?

So short answer: To a degree yes, but unlikely that it is a reliable finding.

True, but they caught up within one year and had the feature. Then they heavily bloated the client, then released a "lite" version which really brought them back on the map. Then facebook came and suddenly it was cooler to communicate non-instantly for some time. :)

It really depends on how you measure the value of the contributions. Simply looking for lines of code (LOC)? Looking for LOC but without drivers?

But even if: is every LOC to be seen as equal contribution?

Because testing everybody is no big deal.

I think this is also a very rational move. IDEs and tooling is just participial about actual performance of said software, but also about the culture and community it is embedded in. So going for a long term commitment to open source is a strong signal to users and other stakeholders.

Bringing up Gitlab in this regard is a very good point, especially after MS bought Github..

Things are not always black and / or white. The rule of law, in reality, is rather a continuous variable then a binary one.

And even if: A kill switch can simply hide / erase some things and present some weak evidence..

Had that car last week for a test drive and I really wondered when the cost of less air drag (while using more energy for display and processing) will break even. The camera mount is quite big..

Yes, in Denmark you buy a car and pay the same amount in taxes, and then you pay half of that in taxes on top ;)

There is an exception for cars under 10.000, I think there the taxes are around 95%.

For that reason, Denmark is full of super small cars and well serviced bigger and older cars.

Yet, I guess the author now moved to Switzerland, at least the number plate does not look danish at all.

Finally I see someone realizing the "hovering" seats in a train. Anytime I sit in a train, I wonder why they have not done so for the last 30 years or so.

While they state its for luggage and design purposes, I just think: Its so much easier to clean. Even automatically ;)

Speaking of logic: Trains are way more energy efficient than planes. So even with the same "strict carbon tax", it would be cheaper to go by train and better for the environment alas.

If it works, the butterfly keyboard is a very good keyboard. For a laptop. It boils down all the nerd arguments about why the IBM model M is still the best keyboard in the world (tactile and overall feedback) and brings it to the modern era. Combined with the restriction of mobile devices. Perfect. If it would not break down so easily.

A friend of mine owned an Audi from 2003 that had a solar-covered sun roof feeding into the AC. On sunny days it would at least exchange enough air while parking so turning on the AC for small trips was not necessary. Just brilliant!

Jupyter Notebooks / JupyterHub and BinderHub are, in my humble opinion, most relevant for the future of education, as a tool for teaching, for science, as a format for replication, and for everything in between, see the data sciences. And even more!

Conda is "just" a distribution and package management tool. As a beginner it is a good idea to just go with it, but installing all packages using pip is just as good, in the end. For reference: https://www.xkcd.com/1987/