On a desktop, systemd and firewalld make sense, because one might have an Ethernet connection that is in a trusted zone, a Wi-Fi adapter that is on a public (untrusted) zone, and so on. Plus, the parallel startup of systemd makes booting a lot faster.

For a server, one wants reliability and security above all. One reason why IBM still obtains boku bucks is because AIX 7.1 still runs applications written for 3.2.5. It might require some compatibility programs to be installed, but if one wanted to run FrameMaker or WordPerfect under Motif, they still can, assuming a graphics card present.

Server-side, it doesn't matter if things start in series. Things need to work properly and be coded for maximum security and reliability.

systemd is the iTunes of the Linux world. It does so much in userland, that a bug in that can mean disaster, or a series of disasters similar to the tons of sendmail holes found in the early to mid 1990s. While it does improve some things, having a large, monolithic package handle so much of userland can mean big trouble [1].

My personal take: systemd is a leap forward. But, for something this crucial to infrastructure, with so many moving parts and so many different interactions between them, this really needs to run through a bug stomping session. Maybe Facebook would torture-test it like they are doing btrfs so that virtually all the major bugs get squashed sooner, rather than later. Even better might be a formal code audit on it (a la TrueCrypt) to find and squash anything that could cause the next Shellshock or RTM worm in coming years.

The one thing that has kept the epic fails out of UNIX is the fact that the OS is made out of a lot of little subsystems. Replace bash with busybox, not that many programs would notice. Replace /bin/yes with busybox's yes... who cares. However, systemd breaks this philosophy. If something breaks, I can't just rename the binary, link in the busybox equivalent, and call it done. I'm dead in the water until a patch comes out, and since this is a subsystem that completely controls the userland environment, this is worrisome when it comes to production critical items.

[1]: Ironic how this is similar to what Tanenbaum said about the Linux kernel.

Hardly. This guy comes up with something cool, and I'm wary of the claim that it will somehow overthrow the existing system, mostly because to informed observers, the current system isn't actually unreasonable (mostly, anyway). There are good reasons behind all of the seemingly-insane details, but they're not as obvious as "some kid is missing a hand".

In fact, I actually have to give quite a bit of credit to the designer of this particular device. On his website, he's not encouraging kids to try the thing or making any claims that it's something particularly special. Rather, he's asking for help from experts to refine the design and turn it into something that is fully-tested and documented. If he can do that and still keep it printable (by end users or even trained technicians... either would be a help), then we'll have a real boon to the state of the art.

I wish him the best of luck, but I also recognize that the obstacles he faces are a bit more realistic (and i daresay more difficult) than fighting a conspiracy theory.

And I can tell you that nerve damage (especially around the fingertips) is important to glove manufacturers, especially concerning sporting gloves, where the risk of such damage is high with or without gloves.

As for watchbands, I actually do know a few people who've had allergic reactions to watchbands of various kinds, starting with myself. I can't wear a gold watch, because after a few hours my wrist turns red, and after an evening of wearing it my lower arm is covered in small red bumps. I have a lesser reaction to my gold wedding ring, but I've never bothered finding out exactly which part of the alloy it is that I'm allergic to. In discussions with others, I've met folks allergic to plastic and cloth watchbands as well as metals, some of whose allergies didn't show up until after months of use.

That's what insurance is for. Sure, it's a slim chance that I'll ever need a $40,000 medical device, but that's why I pay into the pool. If I ever do need it and don't have the cash lying around, my insurance provider does. If I never do need it, then my premiums went mostly to somebody else in the pool who did.

No, I'm saying that the cheapest options present more risks that have not been mitigated. I have no problem informing people of the risks they face, and I sincerely hope that a doctor would inform his patients of the risk associated with any treatment, regardless of the cost.

For a cane, it's a different matter. Canes typically do not have prolonged contact with the wearer and their well-studied risks do not often cause long-term problems once the adjustments have been made.

...and what is the actual risk of that slippery slope? Certainly it's nowhere near probable enough that we'd need to regulate clothing as tightly as medical devices. If you're working with high-energy devices, however, the risk posed by clothing is far greater. I don't recall exactly which jurisdiction requires it, but I know that every piece of clothing worn at my local nuclear plant must be cotton. Cotton burns, while synthetic fibers usually melt. Though often cheaper, synthetic clothes increase the damage from accidents enough to warrant that small amount of regulation.

By that time, the damage may already be permanent. That's one of the things that research would study before handing it off to an unsuspecting patient.

"Functional" prostheses are available for far less than $40,000, and typically are used temporarily while a primary device is being built or repaired.

....until they're dead because the airbag was stripped for resale and the seat belts were worn out.

In other words, you have no idea what a risk analysis is, but you follow the hacker mentality in thinking that you can do anything if you have the raw material and a tool to work it, without the need for actual expertise. As long as the stated objective is met, that's good enough, right?

Medicine doesn't work that way. Medicine (ideally) isn't about just meeting the primary target, but about improving someone's health. Sometimes, that means doing nothing, and occasionally it even means letting people die rather than making the rest of their life miserable. Throughout the process, every decision is based on risk. Every drug has a risk of side effects, every test has a risk of being erroneous, and even if a doctor performs to the best standards available, there is always a risk that their patient will die.

That's why we have the FDA. That's why we run clinical trials. That's why medicine costs so damned much, because someone has to do the research and find out what the risks are, before asking patients to commit their well-being to a new device.

So can you tell me what the long-term effects of wearing this $45 printed device are?

Is it weighted such that it pulls muscles awkwardly, causing pain after a few months of continuous use? Does the constant contact with skin cause any nerve damage? If worn during physical activity, does it create an additional risk of shattering or otherwise injuring the wearer or others?

Can you show test results indicating otherwise, even when the user may not have it attached properly? What resources are available so the user can be certain they're properly fitting the device?

Approved medical devices are expensive because they meet all applicable regulations, and have documentation to prove it. They've been reviewed and tested by experts in the field, who understand exactly what subtle problems to look for that are likely to cause harmful effects in the future. One of the primary principles of medicine is to do no harm. Can you assure patients that this 3D-printed model will be harmless?

Yes, you can buy a beat-up used car for $500. It will still accomplish the obvious goal of transporting you from point A to point B, but it's not going to be as good in the long run as a more expensive one.

Enabling the video camera or microphone won't actually help. You'd need both to determine if the user was actually using their phone, and the processing cost needed to perform that kind of recognition on a large scale would be so ridiculously expensive that it would undermine any additional benefit from the research.

Statistically, a user waiting 60 seconds before searching is uninteresting. It's an outlier, so the developers really don't care what happened. Far more useful would be an observation that 75% of users use the center enter key to submit queries, 20% use the mouse, and 5% use the enter key on the numeric keypad, combined with an observation that 80% of mouse users move the cursor around after a period of inactivity before clicking. To a design team, that means that the users' attention has shifted to typing, and they've forgotten where the mouse is. Perhaps the mouse should highlight in some way when it first moves...

Similarly, the actual content of searches doesn't matter from a UI perspective. If you're having trouble searching for something, it doesn't matter if you're looking for instructions to knit a sweater for a kitten, or the mixture used in the Oklahoma City bombing. On the other hand, the exact search text is useful to the folks developing the search engine, so they can put the most relevant results at the top of the list. Of course, the search engine team doesn't care about how long it takes the user to find their mouse cursor.

This leads to one of the most entertaining aspects of the whole privacy debate. Gathering data is easy, but proper anonymizing is hard. Practically speaking, the analysis of the gathered data is often easier than ensuring that data is anonymous. For example, there are certain combinations of ZIP code and state that identify as few as 30 people within the continental United States, so any data set that includes both ZIP code and state is probably not sufficiently anonymous. It's far easier to simply collect only what's needed for a particular team, and make sure nothing else can be connected to that record. One database records that somebody searched for "geriatric german grandmas spanking spanish men", and another knows that user submitted a search with a mouse, and perhaps another knows that the user is located in western Iowa. With no way to connect the records, the business need is fulfilled and the user's privacy is effectively safe... but the legal disclosure will still simply say that the company collects all those things, stirring up a nice panic.

For me, it isn't the Ethernet port, but the Kensington lock slot. It would be nice to be able to tie down a laptop when not in use, so it doesn't have to be in a rental car in a seedy area of town. Bonus points for a mechanism that deters opening if the lock slot is in use, similar to what the old IBM Thinkpads had.

The certificates would be carried with the cryptographic token. If more info is needed, the old fashioned way of hitting queries is always still there.

The goal is to give people/companies just the info they need to be compliant... and nothing more.

Nail. Head. Hit. I don't want yet another Windows Tablet PC. I want a tablet, but with a docking connector where I can put the tablet in a stand (preferably a stand that has some type of locking mechanism so I can physically lock the tablet down [1].) Of course, a lightweight dock/port replicator would be nice as well, so one could use the laptop as a monitor and a BT keyboard/mouse, and the replicator would give access to USB ports and whatnot.

[1]: It is too bulky, but I'd say the PowerBook Duo dock was one of the absolute best designed docks out there. The laptop was closed and was inserted like a large VCR tape, and locking it was trivial (since it used an active motor to dock/undock.) Maybe something similar for a tablet.

When I saw the iPad, I was assuming it would be the top tier tool for music production, with the ability to handle a lot of virtual sliders. However, in a lot of cases, it only can act as an interface. Can it run ProTools with all the extensions, as well as physically handle the license dongle that some stuff has? Not really. iOS keeps the apps so far away from the device's facilities that a musical application as high end as ProTools or Logic Pro would not be usable.

For music production, a hybrid tablet would be great, especially with Thunderbolt as a way to attach hardware cards. I can see a mini studio that would configured around a device like this, where the device resides in a horizontal cradle and can function as a real time mixer, synth, DAW, and other realtime tasks.

Technically, it sits on a Mach/XNU kernel, with a BSD userland.

If you want a kernel that has an unbroken heritage, the only mainstream OS out there that would have that would be Solaris, which was formerly a BSD kernel, but switched to a AT&T SVR4 kernel. AIX also started out from AT&T code, but went with an odd mix of BSD and AT&T userland items.

All and all, kernel heritage is one thing, but consider the application first. Would someone use QNX for a large-scale database cluster? Not really. Would one use AIX for a realtime microcontroller that has to check a sail switch every 500 ms, and then turn a valve off to a propane line if the sail switch shows not enough air? Not really. There are a lot of UNIX variants (and there were far more in the past... even Dell had their own SVR4 UNIX), so choose the best tool for the job.

AFAIK, the jury is out on that fact. SSDs -tend- to be more predictive due to how they wear out. However, I've not seen any definite comparisons that state that a SSD will have a life longer than a HDD.

There is one limiting factor with SSDs: Once the electrons escape the gates, that's it. No recovery is possible unlike HDDs which the magnetic domains can be present indefinitely. So, as an archiving medium where data is stashed, it isn't very good, unless the media is constantly checked and the data moved periodically.

The a good thing to do with an iMac would be a decent SSD... as well as an external drive appliance with RAID 1, or a volume with software RAID that is similar.

Back when BattleBots was the thing, I considered building one... but you hit on all the major problems, which were only slightly less intractable back then.

To have any chance of making it past preliminary trials, the robot would have to be somewhat successful. That ruled out most of my creative designs. Even making a boring spinner still would have cost a hefty portion of my paycheck, and the required workspace wouldn't have fit comfortably in my apartment. Transporting the thing would have presented more logistics challenges, and even if I'd solved those, the time and hassle to build something just to be torn apart was a pretty steep expense.

Fighting robots is not an everyman's sport. It's a more modern fox hunt.

How will it be leaked, is the question. Usable energy is money, pure and simple, and a disruption will get people with trillions of dollars at their disposal to hide the info, especially anyone in any energy industry. Someone who doesn't get it out far and wide will be 86-ed quickly, similar to the guy back in the Roman times who discovered aluminum, and was promptly killed for it, making a metal too good for mankind to have.

I'd probably say, it would be impossible, once the device gets past the first person. Someone comes up with a working free energy [1] source, as soon as they show it to someone, the inventor is pretty much dead.

[1]: Realistically working... like in the kilowatt to megawatt range. Some gewgaw powering a millivolt LED for a few seconds doesn't count.

A patent will just be violated, and completely ignored. Keeping it secret is the way to go, similar to Heinlein's Shipstones. Place a tamper-resistant box at the client's location, set a meter to charge by the watt-hour, and be done with it. Someone tries breaking into the box, it completely obliterates anything inside showing how it works, or just does a big kaboom, Outer Limits, "Final Exam" style.

On a large scale, build it right on top of a natural gas well. Even though the well is completely empty, nobody will know that and power is power. Done right, one can just use an electric resistance heater to blow hot air out a smokestack so it looks like some combustion is happening. Another option is to use a decommissioned nuclear reactor, pump out some heat to make it look like something is going on, and nobody would even know or care that the electricity came from atoms squeezed together as opposed to blown apart.