Recent Comments

Probably not extinction, but It wouldn't be good. If a good chunk of our modern electrical grid/ telecom infrastructure was destroyed some deaths would occur. Its really tough to predict how many. The increase in radiation would also likely spike cancer rates, leading to more deaths. But probably not mass die offs of the human race.

Ext2 is more compatible than you'd expect. There are 3rd party drivers for Windows and OS X that work very well. The problem is, Windows won't read anything but the first partition on a USB drive, so you can't have a small FAT partition with the drivers, and the rest Ext2. Some firmware tricks could do it, but no companies seem to care about the patent licenses and other problems with FAT32/exFAT, so I end up carrying around two USB thumb drives.

Plus, a surprising number of embedded systems are Linux based, and the developers just don't disable Ext2, so it sometimes works even when not mentioned anywhere.

It would just take a couple digital camera manufacturers to adopt Ext2 (instead of exFAT) and include it in their drivers, and Google making Ext2 the default for Android. Then Microsoft and Apple would quickly be forced to include it by default to avoid being left out and seen as OSes that don't "just work" with accessories.

A weakening magnetic field could interrupt power grids and radio communication,

This does not sound good. I saw once a documentation what might happen when a huge EMP destroys the infrastructure of the modern western world. In the documentation one possible source of such an EMP was an extra-terrestrial gamma-ray burst. So compared to such an event, how does the effects measure to the weakening magnetic field? Might be that such a polarity change never before caused a serious catastrophe. Before the mid-20th century it most likely would not have been much of a problem either. However, the predicted gamma-ray burst EMP scenarios for today's tech/power dependencies came quite close to an extinction level event for humans.

subterranian cable run to my home, higher reliability, and likely higher future bandwidth.

If it would save me the $600 construction fee, I'd certainly be happy to try the wireless, so long as it doesn't glitch too much in bad weather.

And with FIOS using 32-way splitters, wireless could certainly beat it on speed.

My satelite TV suffered from rain fade during storms. Did yours?

No, but then I was starting-off with around 95% signal strength. In fact it took about 4in. of snow collecting on the dish to finally cause a signal outage.

Poor buggers way up in Alaska need a large dish to pickup anything even on a good day...

I would guess that the wobbles more easily explainable by the satellite's low mass. Less mass would mean less force to press the core into a spherical shape.

I've used microwave..that was the licensed frequency stuff that I've used in the past. Now granted it was a different ballpark of equipment than someone using high frequency trading might use so not exactly apples to apples. But what caliber equipment would google use to deliver internet service to the home? I'm betting closer to the stuff I've used than the stuff wallstreet uses. Keep in mind in a point to multipoint environment(unless you're using a satelite) you're not using a directional antenna at both end, and may not even have line of sight--So you're dealing with reflection of signal and all the fun stuff that happens with that.

It seems (http://us.aviatnetworks.com/solutions/low-latency-microwave/) microwave propogates faster than light through a fiber optic cable. Cool. So there's certainly a latency advantage.

But if given a choice between fiber to the home, and microwave to my home I'd still choose fiber, even at a higher price. While I may suffer 2ms in latency, I'll gladly take that in return for subterranian cable run to my home, higher reliability, and likely higher future bandwidth.

My satelite TV suffered from rain fade during storms. Did yours? Also keep in mind that's a direct line of sight link.

The world is sorely lacking a sane removable file system. FAT has tons of technical and political problems and yet it remains the only filesystem that is even remotely feasable for cross platform removable disks.

• Small single file size limit (4.2 GB)
• Small total drive capacity
• Microsoft's long filename patents
• Non-free extensions that only work with Microsoft operating systems

If you live completely in Linux, you can format your drives as ext4. But good luck trying to get Windows or Macs to see it.
If you live completely in Apple-land, you can format your drives as HFS+. But good luck trying to get either Windows or Linux to see it.
If you live completely in Windows, you can format your drives as NTFS. Linux does have multiple competing options of supporting NTFS, but they all basically suck (read only, disables journal, etc.)

Case in point:
My car's stereo system has a USB port that you stick a thumb drive into to play mp3s. Guess what the only supported file system is?
New smart TVs have USB ports that you can stick a thumb drive into to play MP4s, MKVs, and AVIs. Guess what the only supported file system is?
My digital camera uses an SD card to store photos and videos that I take. Guess what the only supported file system is?

Repeatedly using a label tends to (a) diminish the impact of the label and (b) diminish your own argument. All I read there was "trolls, trolls, trolls, trolls". I get it, he thinks those against systemd on principle have their allegiances and logic misplaced. Terrible way to tell it though.

Well at least we can all agree about diets then. :)

Over on Metafilter I found out that Steve "Buy My Unnecessary 1992 Disk Repair Program For Your 2 TB SATA Drive Today" Gibson is now pushing a diet based on eating primarily fat and forcing your body into a steady state of ketosis (often a feature of Atkins style protein diets).

I was perfectly fine with Mehmet Oz doing a Dr. Oprah show, until I found out he's also a pusher of reiki and various other nuttery because his wife is into it.

Cut the scientists some slag. They need to convince people to give research funding, who cannot distinguish Harvard from Hogwarts.

Good tip, thanks for it. I use aldiko (free version) and moon reader, both on Android, and both quite acceptable. I might try this - looks also useful for reading epubs on a Linux desktop (I currently use calibre for that. Calibre was roundly trashed at this previous |. article)

It was a response to evilviper and his or her reference to a prior thread, in which the insect story was very much positioned by one of the scientists as if it had immediate analogs to human reproduction.

Trying to convince people not to too-firmly base their conclusions on some currently accepted theories where the supporting evidence is weak or there are known unresolved problems.

Not sure if I agree here. I have quite a good scientific education. However, in 99.99% of all scientific fields I am just layman. All people are. Nowadays nobody can have a complete overview over science. Not even a complete overview in once specific field, e.g. physics. So you have to go with the masses = currently accepted theories. And this is fine as long as one has a base knowledge how science works: You develop a hypothesis. You try to find evidence, which supports your hypothesis. And most importantly you also try to find evidence, which disproves your hypothesis. If something disproves your hypothesis, you drop it immediately, or try to adjust it so that there is no contradiction. This way you can develop your theory. Weak evidence? As long no contradicting evidence not a real problem just a reason for more research. Known unresolved problems? Does not necessarily devalue your theory. Might be that it can be extended. DNA inheritance is not wrong just because there also are epigenetic effects.

I am similarly cautious about theories on dark matter,

I am not. It is the currently accepted theory. It does not contradict anything else I learned. I am not able to disprove it, or do otherwise substantial work on this field. So I accept dark matter as what it currently is: An attempt to explain certain observations. If anyone comes with a better explanation... I'd immediately drop dark matter. Give me enough evidence I'd forgo everything I learned. Give me enough evidence, and I 'believe' in unicorns and magic.

Btw... to be exact: The existence of dark matter is currently no theory, but only a hypothesis. To become a theory it needs evidence for its existence beyond being a pure mathematical trick to explain otherwise unexplainable observed gravitational effects.
Furthermore in science you cannot say 'it is only a theory'. There is nothing 'higher' than a theory in science. Theory of relativity (general or special) I am not sure there is a theory, which has been so thoroughly tested. Probably thousands of experiments, which confirm the theory of relativity. And it is still a theory... an will ever be... unless someone proves it wrong.

It's more of a nuisance with nutritional or diet theory-of-the-week,

Nutritional or diet theory is mostly neither a hypothesis nor a theory.... the best term to describe most of this field is 'religion'.

Actually, microwave links are ridiculously reliable. Before fiber optics, before communications satellites, most long-distance phone calls and TV programming went from microwave tower to microwave tower, all the way across the country, as needed.

Today, microwave links from NY (or NJ) to Chicago are competing with and replacing fiber-optics, particularly for high-speed trading purposes, due to the lower latency.

http://www.heraldonline.com/2014/09/23/6352231/innovator-windy-apple-announces.html

http://www.computerworld.com/article/2493686/financial-it/microwave-vies-with-fiber-for-high-frequency-trading.html

http://www.bloomberg.com/news/2014-07-15/wall-street-grabs-nato-towers-in-traders-speed-of-light-quest.html

You shouldn't confuse your experiences with companies that have "wireless" in their names, that are possibly operating very lean and selling sub-par service, with the underlying technology. I think we can all agree that satellite TV can be pretty darn reliable... if it wasn't, your OTA or cable TV wouldn't work, either, because they get their network programming feed via satellite. Ditto for, say, OTA TV, FM radio, etc. There's no reason local wireless from the end of the block couldn't be every bit as reliable.

It's a debate I keep having... Trying to convince people not to too-firmly base their conclusions on some currently accepted theories where the supporting evidence is weak or there are known unresolved problems. Just because nobody has disproven theory X yet, doesn't mean it's a good idea to go out and start bloodletting sick patients...

DNA/genomics was particularly solid, but had some red-flags in the form of obvious outward differences of DNA-identical twins, which epigenetics is now helping to resolve.

http://multiples.about.com/od/funfacts/a/Identical-Twins-And-Dna.htm

I am similarly cautious about theories on dark matter, most conclusions drawn from the rather patchy fossil record, etc.

It's more of a nuisance with nutritional or diet theory-of-the-week, and generally people not well-informed enough to see Dr. Oz and his ilk as the bald-faced lying flim-flam artists they are.

https://en.wikipedia.org/wiki/Mehmet_Oz#Scientific_validity

Oops: internet outage + "click submit again" = dupe comment. Sorry!

I had to Google "eloi" and came up with this, courtesy of Wikipedia:

The Eloi are one of the two post-human races in H. G. Wells's 1895 novel The Time Machine.

Nice one. Too bad your graphic didn't embed, though. Having run a forum (gotonicaragua.com) that had bots embedding graphic ads into their spam posts, I can see why it isnt permitted though!

Interesting post here:
http://www.vitavonni.de/blog/201410/2014102101-avoiding-systemd.html

Guy makes a couple of good points. Here are two clips:

Avoiding systemd isn't hard

Don't listen to trolls. They lie.
Debian was and continues to be about choice. Previously, you could configure Debian to use other init systems, and you can continue to do so in the future.
In fact, with wheezy, sysvinit was essential. In the words of trolls, Debian "forced" you to install SysV init!
With jessie, it will become easier to choose the init system, because neither init system is essential now. Instead, there is an essential meta-package "init", which requires you to install one of systemd-sysv | sysvinit-core | upstart. In other words, you have more choice than ever before.
Again: don't listen to trolls.

He goes on to point out, however, that you will get systemd installed if you use the gdm Gnome login manager or any part of the Gnome desktop, so for the moment, not only is systemd optional, but you can avoid it if you also avoid Gnome.

On a server, there shouldn't be any component actually depending on systemd at all. systemd is mostly a GNOME-desktop thing as of now.
As you can see, the trolls are totally blaming the wrong people, for the wrong reasons... and in fact, the trolls make up false claims (as a fact, systemd-shim was updated on Oct 14). Stop listening to trolls, please.
If you find a bug - a package that needlessly depends on systemd, or a good way to remove some dependency e.g. via dynamic linking, please contribute a patch upstream and file a bug. Solve problems at the package/bug level, instead of wasting time doing hate speeches.

I don't have any comment on the facts, since there isn't very much more to say. But I will point out, that is one awesome looking moon! ("That's no moon ...")

100 internet points for the man who used logic plus science to make an educated guess, and got it right: sickle cell is a genetic mutation which enables the victim to get malaria and have less severe symptoms, increasing the carrier's chances of survival and messing with the parasite's life cycle. Evolution is a bitch, though. If two people with the gene have 4 kids, one of those kids will likely get the disease associated with it, sickle cell anemia, and be more vulnerable to malaria. It's a net win for people living in a malarial area, but the double-carriers still get a raw deal.

Actually I used the word 'controversial'. My story got somewhat improved by the editor (thank you), so it got dropped.

But where do you see that someone want to convince people that insect genetics has direct bearing on
humans? The only thing that was said was: 'A new and unexpected mechanism in the reproduction process
of insects has been discovered. Let's see if it or something similar exists for other species, too'. Epigenetic
effects have already been proven to exist for animals and even humans. Epigenetic effects provide a plausible
mechanism for what was discovered with the insects. So IMHO it is not far fetched to at least look, if something similar
exists outside the insect kingdom, too.

He didn't use the word "controversial" so there's no real need to quote it. Boy you're really hung up on convincing people that insect genetics has direct bearing on humans. It's interesting science, but that's all it is.

Okay I finally RTFA. They give up on the idea of common carrier immediately, without a fight, and their reasons make NO SENSE AT ALL. They compare how "easy" it was to require common carrier for telco traffic but somehow the Internet is "vastly more complicated". Their argument is bullshit.

"Congress enacted Title II in 1934 primarily to regulate telephone service. Telephones of that era delivered exactly one functionality: real-time voice transmission. Non-discrimination meant that everybody got a dial tone on equal terms. That was easy to regulate. Enforcement was easy, too, since one company handled local service in nearly every city and town, and was also the country's only long-distance provider.

The Internet is vastly more complicated, with astronomical numbers of providers and services. A simple rule saying nothing more than ISPs "shall not discriminate" would be meaningless. An ISP's capacity is, after all, finite. At peak times it may not be able to accommodate 100% of all potential content - email, Facebook posts, Netflix video, VoIP calls, people working from home, casual browsing. At those times, some discrimination must necessarily occur in allotting access to providers. The question, then, is how to ensure that the discrimination is "fair". An effective non-discrimination rule would give an ISP managing a traffic overload clear guidance on which bits to send on and which to hold back in every possible situation. More than that, a proper rule would let the ISP program in algorithms that make these decisions automatically, on the fly."

What, there was never a capacity limit on the ability of a local CO switch to handle voice traffic? Everybody gets a dial tone, no matter how many highrises suddenly appear? Let's see, where would that capacity come from... from the provider (AT&T in that scenario) PROPERLY MAINTAINING THEIR NETWORK. The ISP business is exactly NO different. They should keep their network maintained and upgraded to allow the traffic that they are actively selling to business and consumer. "An ISP's capacity is, after all, finite." As if 1930 AT&T's capacity was magically infinite.

What a load of hogwash. The entire article is a waste of time. You can tell it was written by lawyers and not technologists.

Even their "solution" reeks of magic non-techie thinking. Oh, the miracle of competition will suddenly mean that congested hybrid coax-fiber infrastructure and leased lines will be able to deliver higher capacity just by changing a bill-to address. Have these people ever SEEN a cable modem?? You're on Time Warner's cable infrastructure, you get the capacity of Time Warner's cable infrastructure, no matter who is branding it and collecting the checks.

Thank you. This news just made my day.
So little do we know...

Feel free to continue to use what works for you.

It's not just a tiny number of organizations though. More and more organizations are deploying virtual machines and containers. In a more dynamic server system like this, the old way of doing things shows its limits. I have written my own restart scripts on many occasions, but it sucks and I'm tired of the hacks. Even in a small organization, I've had race conditions with NFS, openldap, and network connectivity stop booting in its tracks on many occasions. I used to change inittab to start a getty right at the beginning of the init process so I could log in and kick things when necessary (requiring physical access to the server, which sucked big time, but virtualized consoles improved that dramatically). So even for small organizations of one or two servers, the init script system does have problems still.

That this stuff is happening - in the USA, at least - despite a culture increasingly hostile to the "educated elite" and whatever other impolite names the likes of Sarah Palin came up with for people who like science -

Who cares what the Eloi think?
http://s25.postimg.org/yse3ijo4r/b5u8_2l.gif

I've been paying closer attention to other fields of science recently, and wasn't aware of any of this stuff, so I find it endlessly fascinating. Every generation, I think, is willing to laugh at the false theories that previous generations accepted as fact before eventually finding there was a better explanation, without taking the time to wonder how many of their "facts" will eventually be overturned by future scientists looking for more accurate explanations.

That this stuff is happening - in the USA, at least - despite a culture increasingly hostile to the "educated elite" and whatever other impolite names the likes of Sarah Palin came up with for people who like science - is even more fascinating. Kudos to the boffins!

Maybe you should read up a bit on systemd before you post this kind of untrue stuff, or maybe run it and see just what it does. Systemd does *not* cram everything in pid 1! To say it does is false, plain and simple. You'd know this if you took some time to examine a system that runs systemd. It also does not require a reboot for most updates to systemd. Everything that can possibly be done outside of pid 1 is done outside of pid 1. If you do an update, the package manager issues systemctl daemon-reexec and all the systemd subsystems restart with the new binaries. If something was changed in pid 1 then, like upstart or system v, a reboot is required. But that just doesn't happen very often. I've installed maybe 2 updates to the systemd packages on RHEL 7, and did not need a reboot.

I don't mind actual arguments being made against systemd, but I rarely find any in these discussions! Just mumblings about philosophy and theoretical problems, completely ignoring the very real and ongoing problems with system v and the like, which are impacting a lot of sysadmins. Which is pretty crazy. We've had maybe a half dozen stories on it in the last month on the other sites, and no one can say anything other than FUD it appears.

I'm not arguing for systemd so much as taking issue with the way people are spreading this sort of misinformation, and taking issue with the idea that system v wasn't broken (it is broken, especially with modern, hot-plugged server hardware).

Great, so let's accept your position that the tiny number of organizations who rely on "hundreds" of like servers and can't be bothered to write their own restart scripts are to benefit from an all-encompassing process controller like systemd.

Okay, so why burden the rest of the world with the same monstrosity when they (a) DON'T need it and (b) might actually prefer to prevent process crashes and maintain their own systems to restart them?

Also, I note that an actual real-live registered user chimed in and you are now powerless to complain about ACs this time. :)

I use my tablet (Galaxy Note 10.2) almost only to read eBooks (>90%). I always read in landscape mode if the format is epub. This way I have two pages open in almost the same size as when I read a normal paper pocket book. Of course, technical literature in pdf format I usually read in portrait mode. Whatever works best.

Btw...when I bought my tablet a few years ago, I tested quite a few readers. At that time cool reader was the best I found:
http://coolreader.org/e-index.htm
Maybe in the meantime a better one was released, but I am still happy enough with this one that I feel no pressure to search for
alternatives. One of the best features: It is open source.

I find this very interesting. I wonder when this is commercially used. Gene technology is widely frowned upon. Mostly by people with no sufficient knowledge to be able to have an informed stand is this matter, and mostly for the totally wrong reasons. So, what is epigenetics able to do? Could it be possible to 'gas' seeds with some chemicals and make them this way more resistant against pests? Fine tune the fat/meat ratio in a pig? Without changing the DNA itself? Could be a billion $ market.

There is no replacement for Fiber. There's things people try to do to cheap out and give a inferior product, such as brighthouse's "hybrid fiber coax" network. I've used wireless before, even licensed wireless links: They absolutely suck compared to fiber. I'm not talking about bandwidth I'm just talking about you know--actually working all the time.

Wireless does incredibly crazy stuff(depending on spectrum). It bounces of things, multiple copies of information arrive from different paths, rain screws with it, microwaves screw with it, winds screw with the antennas, etc.

If google fiber wants to change itself from the absolute premier internet provider into something only a backwater coal town with dialup would want, feel free to change to wireless. I'll stick with my FTTH network.

I believe the only thing "controversial" was that one AC just patently refused to accept that epigenetics exists, and even the most carefully qualified possibility it could possibly affect humans.

It's not too surprising, as it's a new enough field that probably every one of us were taught in school about strict DNA inheritance, with no room for other mechanisms like the emerging field of epigenetics.

http://www.pbs.org/wgbh/nova/body/epigenetics.html

Even experts have a hard time accepting it:

"Genetics. It turned out to be more complicated than we thought." --Laura Hercher

"My instinct is deep skepticism" --Kevin Mitchell

https://answersingenesis.org/genetics/epigenetics/epigenetic-changes-let-mice-inherit-their-fathers-fears/

But the fact that humans experience certain epigenetic effects has been rather firmly proven:

https://en.wikipedia.org/wiki/Epigenetics#Epigenetic_effects_in_humans

It happens to be a fact, and nobody has even attempted to refute it, so apparently everyone agrees, and just wants to pretend they didn't hear it due to how badly it destroys their world-view.

Just lumped-in with SysV, as there isn't a lot of difference there.

Similar here, although even without the keyboard I tend to surf the web, read email etc. in landscape mode. I generally only use portrait mode when I'm reading an eBook (for which I always remove the keyboard and use portrait mode) or for some web sites, or even specific articles, that I think work better that way around. Probably that's just pre-conditioning on my part and that of web site designers; we're still used to the landscape orientation of desktop monitor being the norm, even though some recent web stats imply that most browsing for some sites is now done via phones and tablets.

+1 for the bath

Wrap it in an A4 ziplock bag

Jeez, plus one to that, brother. I actually relocated the modem/router to make sure I could get a signal in the bathroom. Living the good life now!

Strongly agree.

It happens to civilizations and countries and corporations, and it can happen to linux distros too. It's perfectly possible for them to get corrupted from the inside. Humans are at the controls, and no humans are perfect.

Services will and do crash for no reason

Appeal to magic? I think if you give the matter even a tiny bit more thought you will agree that it is never "no reason". It may be an extreme low-runner software bug or a hardware glitch, but those are reasons. We may not always be able to figure out the reason, but it is always there. Heck, even an alpha particle in a memory cell is not "no reason:.

And repetition does not magically make an argument persuasive.

No, upstart is in there, too, and that's about all

Is there some genuine reason for dismissing OpenRC?

Have a look at Australian ISP prices. They all compete in this way. DSL 20/1 for around $50 a month from a range of ISPs.

Now they are competing with each other to install fibre. This is the kind of environment we need to get 90% of Australians on broadband. Pity the NBN is flawed and trying to take everything down with it.

I should say that when a service crash I prefer to have it restarted as soon as possible on another spare server in order to be able to investigate the crash before reusing the server hosting the crashed service.
Therefore, I dunno want systemd.
Boot time are irrelevant as a spare server should always be available and a new spare provisionned as soon as the one is used. (hopefully the crashed machine can be bringed back as spare).

the bathroom

I prefer portrait mode when the tablet is alone. But with the ASUS Transformer, I often have the tablet attached to its keyboard dock - which pretty much locks it into landscape mode.

I'm not so sure. The Nexus 7 is also popular for permanent setups like in-car installs because of its size. The great thing about it is that it fits so many different use cases.

Dude, regenerative braking. Its not rocket science.

Forking Debian would be quite a bit of work, I'll have to agree. Easier would be to take the ALREADY supported sysV init packages, along with other compatible packages that Debian has kept around as optional, and roll a forked installer that still provides Debian as maintained by Debian. This seems like it'd be a bit more of a sane way to handle the undertaking, and avoid splitting the labor pool unnecessarily.

Systemd's existence isn't a problem. Poettering's statement of "Linux is still too fragmented...[and] needs to be streamlined..." (Wikipedia reference due to original source being in German) is however a bit of a worrisome attitude, as it seems to have a monoculture for Linux as the aim such a sentiment espouses. That the rollout of systemd has seen established software losing cross compatibility with minimal benefit is also troubling, and appears to be a sharp move in the wrong direction.

When systemd can learn to play better with others, I think you can expect the uproar to die down a bit. Expecting a community made up of people who go out of their way to use a system that has heretofore been one of the least one-size-fits-all in its philosophy to be thrilled by software (however functional for what it aims to do) that breaks that trend is a bit foolish if you ask me.

Seems like a matter of trying to bring in the masses at the expense of alienating one's already established user base. Perhaps we should go ask Slashdot how well that works.

The idea of there being one init system to rule them all and in the darkness bind them I think has a LOT to do with negative attitudes toward systemd in the first place. If Poettering could spend more time coding and less time actively seeking the destruction of other functional, stable, and good software (before having anything all that stable/functional to replace it with, in many cases), he'd be a lot more welcome in my book. Pulseaudio works great when you need what it has to offer, and while there may be quirks about it I'd like to see done otherwise, well...that's hardly something I'd have much place complaining about not producing something that does what it does to replace it with. What doesn't work great is when you take out an underlying functionality and replace it with something unfinished and broken. When you simultaneously make going back purposely difficult or impossible, it's even more insidious.

I seriously hope he mellows out as he gets older. This "aint monoculture neat?" trend really is getting quite old, and for someone who apparently can make decent software tools FOR THOSE THAT WANT THEM, he really is wasting that on playing this Borg-ish villain role.

In any case, I'll keep using my SysV setup with BSD-style init scripts and pulseaudio and enjoying an open Linux environment while it lasts. Hopefully Debian can figure out a reasonable way to navigate the path ahead with some grace...it'd be a shame to see them dashed on the rocks by this state of affairs.