hacst

Some providers do what imo is a best of both worlds approach here: Every customer has a full fiber run to the PoP, but there they use GPON to save on the active components. The actual fiber is pretty cheap compared to actually bringing it into the ground and that way you retain full flexibility.

That matches my observation on Windows 11. Looking at wireshark while doing the resolution I see the following: Normal queries never use the the suffix. nslookup does, which the resolver answers with NXDOMAIN

Fritz box allows you to capture traffic towards the ISP too and there I never see such a suffix query in either case. So I assume the fritz resolver directly responds to those.

That does not really match the behavior described in the OP but I would be very surprised if the behavior described there wouldn't have lead to a big outcry much earlier.

That's what I tried to capture with the "at the population level" restriction. Plomin was quite clear that for the individual all bets are off so sorry if it sounded like the environment couldn't still screw you over (it definitely can). The part about inter-group differences is worth emphasizing and I should have done so.

That did not reduce my surprise about the things stated in the book though. If statements it contains like that after correcting for genetics "the most important environmental factors, such as our families and schools, account for less than 5 percent of the differences between us in our mental health or how well we did at school" and "Genetics accounts for 50 per cent of psychological differences, not just for mental health and school achievement, but for all psychological traits, from personality to mental abilities" are true, then I definitely consider that as going against "common knowledge" and "common sense". Definitely blew my mind.

Thanks for the links. Totally missed that it was discussed here before.

I recently read the book "Blueprint: How DNA Makes Us Who We Are" by Robert Plomin and it it was quite the eye opener as to how much the nature vs nurture debate seems to have been overtaken by genetic findings.

According to the book genetics seems to be the single biggest factor for a ton of things in our psychology. Including various measures of intelligence and achievement. As a layman the evidence presented seemed quite convincing. The book mostly refers to big studies with thousands of twins. To separate nature and nurture studies can do things like tracking twins given up shortly after birth into two different foster families who never had any contact to their birth parents later (correlated nature), track the development twins in the same family (correlated nurture) and so on.

According to Plomin the effects are huge (especially compared to what usually would be a significant result in a psychology study looking at nurture) and the controversial nature of such findings made the field extremely rigorous to the point where it is hard to imagine these results turning out to be wrong. Apparently it has all been replicated quite a bit by now.

To be honest I found the book quite shocking because it runs counter to so much that is "common knowledge" and "common sense". It basically mostly discards influenceable "nurture" as a defining influence at the population level. Once certain basics are met, relevant environmental influences are mostly too random to control and much of the rest is indirectly caused by "nature". "Nature" basically creates its preferred environment (called "nature of nurture" in the book).

As I am not in the field my ability to verify what the book says is limited. As far as I can tell Plomin is a very well regarded Psychologist and mostly know for his involvement in twin studies. If anyone in the field has better insight on how to evaluate what the book says I would be very interested.

In my mind the even bigger questions is whether it being "cheaper" will stay true once we actually try to scale the use of the technology. As in can we use this to actually produce most of our energy needs and still have the cost benefit?

I'm not an expert on energy production but a study I recently read on this topic for the German "Energiewende" over here did not sound so great [1]. In Germany the renewables we can scale are mostly Wind + PV. However according to the study their seasonal patterns combined with the demand curve seem to make it pretty much impossible to actually expand them meaningfully. Over the year you'll get peak production far outstripping demand (we've already seen negative prices in our markets with current installed capacity) and in the troughs you will need huge production from other energy sources (mostly old coal plants at this point). As these are seasonal patterns this means you would need to store TWh over many months to average that out and make new installed renewable capacity actually useful. The study did some back on the envelope calculations for various options for that but none seemed to come close to achieving that.

If that is true, system wide such nice looking renewable price calculations are preconditioned on a backup structure of conventional plants being on standby to fill in the throughs. As this is highly uneconomical for them the cost for that will still have to be paid by the system. At the same time due to the extreme variability prices will drop to zero during peak times (well...unless you have feed-in tarifs) making cheap production prices still be a loss.

I imagine the calculations might look much more favorable in other locales where demand is better correlated with renewable production but I would be surprised if Germany is the total exception here.

[1] http://www.hanswernersinn.de/dcs/2017%20Buffering%20Volatili...

Afaik you can still configure it this way in Pro and higher skews using the group policy editor. And it's not like they didn't try the "asking the user" thing before. It's just that many users will happily keep dismissing any kind of warning that doesn't immediatly make their computer stop working forever. That's just no longer something that's acceptable for a machine connected to the internet.

They do raise this in their discussion part: "Fourth, vaccines are only active while pathogens are inside hosts, but drugs can remain active in environmental reservoirs [89], suggesting that the strength of selection for resistance may differ for drug and vaccine resistance. However, drug resistance readily evolves even in pathogens that lack environmental life stages such as HIV [8]."

You must not be able to prove who you voted for. Otherwise you are still at risk of retaliation, coercion and can sell your vote. For the very same reasons the government must not be able to know who you voted for either.

Secret paper ballots really deserve more credit. The tech might be old but it gets the essential features right.

There is https://caremad.io/2015/04/a-year-of-pypi-downloads/ which is also based on PyPI downloads and is more recent.

Never considered the sarcasm and social critique in that movie to be all that subtle. "Would you like to know more?" (https://www.youtube.com/watch?v=faFuaYA-daw). Loved the movie for it.

Could never get over how different the book felt though. No hint of what I liked so much in the movie. Basically the complete inverse message of the movie.

Don't think I've ever seen this kind of "same story, inverse message" between a book and a movie before.

I must be missing some communication medium / tracker reference or is the information listed in the code-review really all there is on this change?

Full commit message: "Tweak RNG." Reviewer comment: "LGTM" landed

No reason for the change, no source code comment, no discussion?

I was pretty impressed by the mixed native/python debugging capabilities in PTVS (https://github.com/Microsoft/PTVS/wiki/Mixed-Mode-Debugging). It's really handy to be able to just step into native code and back seamlessly if you project has native components. Not sure if PyCharm offers something comparable.

Optimizing compilers.

At the very top you'll probably have to search very hard to find a sport that is actually good for your health. https://www.youtube.com/watch?v=8XnQyI3Mgbc is a great documentary on what it takes to win gold in the Olympics. These athletes knowingly - and willingly - trade their health for a shot at their goals.

I have actually though about a boiled down client with baked in default configuration before. I wasn't considering business applications though but was thinking about the use in podcasting. Never got started on it though.

We work on Mumble in our spare time and as such developer time is scarce. We try to focus on what we think most users are going to benefit from. There's a long "Would be neat to have X" list ;)

One goal for the future is to make Mumble more modular which should make creating such a specialized client feasible without much extra work.

If someone here is interested in C++ with Qt and wants to contribute to a FOSS project used by nearly 400,000 users monthly we can always use more hands.

Mumble core developer here. Glad to hear Mumble is useful to you and your coworkers. Always nice to get positive feedback.

If anyone has question about the software or the project feel free to ask.

The header compression in HTTP 2.0 isn't based on gzip or something like that. The CRIME attack pretty much killed those approaches dead. It's more akin to differential updates for header during the lifetime of the connection. So if you request a lot of files with fairly similar headers you'll effectively only have to transmit the bulk of the header once while the other request will efficiently re-use the previously transmitted fields.

So to answer your question: Header compression as employed in HTTP 2.0 helps if you do many requests with similar headers on the same connection.

For more information about container shipping I highly recommend this omega tau podcast episode: http://omegataupodcast.net/2014/04/146-container-shipping/