Google should do whatever it wants. After all, if I get annoyed enough by Google Chrome, I'll just switch back to Firefox or Opera. Only the ChromeOS/ChromeBook/ChromeBox users may be screwed (because they've made the mistake of locking their hardware to a specific vendor browser).
IE taught us that this kind of thing doesn't happen quickly - web developers _still_ have to deal with IE's buggy rendering, despite good alternatives having been available for 15 years. Ok, IE has got better but it's still not great. Users don't see this stuff as a browser problem - if your website doesn't work right then the users see it as a problem with your website.
Well, it is pretty much like real life, but I'm not sure we want to be teaching kids "this is the crap you can expect fom life" rather than inspiring them to do more.
I did take a slight exception to this though:
But Steven and Brian are also everything frustrating about the tech industry. Steven and Brian represent the tech industry assumption that only men make meaningful contributions.
As far as I can tell from the story, Steven and Brian did nothing wrong at all - clueless Barbie fucks things up and then asks them to fix it, which they do. This bit of the story would probably be pretty similar if you replace Barbie with any clueless person (male or female) who's just infected a bunch of computers with a virus. What were they supposed to do in this situation?
An elite crowd trying to force on everyone else what they think is the right way? Thats one of the many reasons people are against systemd!
The maintainers (you call them "an elite crowd") of some distros have made the decision to use systemd because they think that's the right thing to do - someone has to make the decision, and if not the maintainers, who? Or would you prefer that the maintainers decide to do something that they think isn't right?
No one is forcing anyone to use systemd - the source is there for anyone to use as they see fit; Some distros have decided that systemd is the right way to go, some have decided to use other inits, you can either choose the distro (from a wide selection) that suits your purposes the most, or you can even make your own, no one is forcing you to use one particular distro.
Note: I don't really have any opinions about systemd, I currently use Fedora and it seems to work ok, but if I have problems then I can switch distros.
One thing I don't understand is how in the hell it is considered ok to have this in Debian STABLE? Maybe, in Fedora or OpenSuse but Debian stable???!
Why not Debian Stable? Red Hat Enterprise Linux uses systemd, so it must be good enough for enterprise use, so why it it not good enough for Debian Stable?
"systemd does the right thing by stopping normal boot and just boot into a safe, minimal shell. A quick glace in the log file (journal) will instantly tell you (using red letters for emphasis) that fstab is broken in such and such a way. A quick edit with Vim can then solve the problem." - did you miss these lines in his comment? Just how "far" is "far enough" ?
Well that would depend... If its your desktop machine then popping a shell on the screen would probably work(*). If it's a headless networked device then you're going to need the NICs brought up and sshd started.
(*) This isn't especially user friendly though... how about firing X up and having a nice GUI thing to fix the problem?
This kind of tight coupling is unheard of in Linux history.
Not true at all - stuff has been tightly coupled plenty of times in the past. Lots of stuff is very tightly coupled with udev these days, for example. And whilst I will agree that tight coupling is bad, its sometimes hard to see how it could be avoided.
The use of System V init allowed Linux to be comfortablef for UNIX admins looking for a less expensive or more widely installable solution, and the end of the use of System V init means that Linux is starting to head away from the UNIX operating systems.
Linux has been heading away from Unix systems for a long time. As a long-time Linux user, on the odd occasion that I have to deal with the likes of Solaris I find it feels *very* backwards by comparison... It's almost like going back to the 1980s...
What if it was someone attacking your sshd and making it crash when it failed?
By automatically restarting it, you just allow the attacker to continue trying to exploit it.
By automatically restarting it, you don't solve the issue that makes it crashing.
By automatically restarting it, you, most of the time, don't even see it restarted, so really not giving you any way to solve the real problem.
It's not that I don't find process monitoring interesting, it's just that automatically restarting can bring more problems than it solves.
As with any service, the "correct" action upon a crash is probably dependent on what the machine is actually supposed to be doing. Take for example, a dedicated web server - having Apache do down when under attack and not attempt to recover would be bad since the attacker will have successfully caused a denial of service with very little effort. Compare to a private telephone exchange, for example, which is running a web server purely for management purposes - a crashed web server is not a disaster, the whole thing keeps doing its primary job without it, so automatically restarting the crashed web service _may_ not be the best plan.
So I guess the answer here is "it depends" and therefore the administrator should be able to choose either option, so selecting an init system that doesn't support one of the options would be bad.
In the case of sshd, since it is potentially the only way to safely fix a broken server, allowing it to die permanently seems like a bad option to me. A better option would probably be to restart it and firewall off all but a few "safe" IP addresses. That way the administrator can still access the server from one of those IPs and the attacker can't cause any more damage.
A well behaving daemon shouldn't be restarted (except maybe for rereading config files), it should start and stay that way. If it crashes randomly, then you might try to find the bug.
Whilst I agree that you should fix a crashy service rather than restarting it each time it breaks, there are nver the less reasons why you may want to auto-restart the service:
- In the real world, you can't just shut down a service until a bug has been fixed; you need to continue running it as best you can while the problem is being looked into and fixed. So a stop-gap measure may be necessary.
- Whilst you may believe some software to be bug-free, this may not be the case, and in some cases it would be disasterous to discover that thre is a bug by finding a service permanently go down. Far better to restart it and log the error.
- Bits _do_ occasionally get flipped in memory or registers, so software may well occasionally crash through no fault of its own. It is reasonable to have something in place to mitigate this should it ever happen. So yes, I agree, if a service is crashing all the time then it needs to be fixed, but that doesn't mean that you should abandon all possibility of recovering from an unexpected crash.
Comcast offered a $50 cash card if we signed up for internet service with them. We signed up in May, and the card never came. We called and they denied that they ever offered the card. A few more calls later, they agreed that they offered it and said they would send it. It never came. Last month, five months, a final call was made and the card arrived.
Clearly they have a strategy of screwing customers, either through intentional scripting or extreme negligence.
Not just telecomms companies - I'm currently being screwed over by Npower. I was a customer for 18 months, during that time I got a single correct bill and had to spend hours chasing them to get the others corrected. My original contract gave me a discount on the billing anniversary (January) - they never credited the discount, so I raised a dispute in January and they agreed to credit it. Except it never got credited. In the summer I left them as a customer, filed a complaint (about the shiteness of their service and about the unpaid discount) and refused to pay the final bill as it was incorrect. They responded to my dispute, replying to my complaint about the poor service but ignoring my points about the incorrect bill. I re-raised the complaint and got an automated "we'll respond in under 10 days" reply - never got an actual response so I followed it up 10 days later, again no response. The billing department are now threatening to take out a court summons against me for the unpaid (incorrect) bill. The billing department say they can't do anything about the disputed bill or the threatened court action and that I will need to raise a complaint with the complaints department, completely ignoring the fact that I've already tried to raise a complaint several times and the complaints department won't respond.
So now I've sent 50MB of paperwork and telephone recordings to the regulator in the hope that they can beat some sense into Npower before I have to waste yet more time defending myself in court.
I'm left wondering if this is actually incompetence, or if their corporate policy is to conveniently "forget" to pay discounts that were promised a year ago in the hope that most people won't notice.
It's somewhat like buying a last minute airline ticket. If people were unwilling to pay more for a last minute ticket, all tickets would cost more (fine) but it would be impossible (because the airlines would price tickets to insure every seat was sold - or oversold - many hours before wheels up to minimize the risk of a single empty seat) to get a ticket on a commercial airliner to get to mom's bedside 1500 miles away before she expires.
Airline tickets are an interesting one; and other products that have a relatively inflexible supply - i.e. the costs of flying the plane are basically the same whether or not that seat is sold, and similarly you can't sell more seats than you have, so the supply is inflexible. There are two opposing forces at work here: 1. The airline wants to ensure that every seat is sold, since filling a seat at the last minute increases profit, even if it is sold for below cost (the plane is flying anyway, the costs can't be avoided, selling the seat rather than flying with it empty is beneficial no matter how cheaply you sell it). This is going to tend to push the prices down for "last minute" sales as the airline tries to attract sales. 2. The passengers that need to get somewhere at short notice are willing to pay a premium. This is going to push the "last minute" prices up as the airlines cash in on this willingness to pay over the odds.
Figuring out which of these forces wins is certainly a non-trivial exercise.
By increasing prices when demand approaches the absolute maximum supply, consumers will reduce demand quickly (good, since supply can't be increased quickly). When power gets expensive enough, they will shut off rooms, wear more sweaters, turn lights off, instead of cooking a fancy dinner they will nuke something in the microwave and use disposable utensils (or, just wait to wash them until the next day), they will sit around in a single room and talk instead of playing on their computer or watching TV in individual rooms. Demand is extremely elastic, supply is inelastic at the top end. In extreme cases, they will shutdown their entire house (using winter shutdown procedures as needed) and gather in friends and neighbor's houses (perhaps, splitting the cost of the very expensive power during those times).
I think expecting people to monitor electricity prices on a minute by minute basis and change what they are doing _now_ is (largely) not realistic - virtually no one is going to look at the electricity price before deciding to put the TV on, for example. What is realistic is getting people into a routine - if people know that it's always cheaper for them to put the dish washer / washing machine / whatever on over night, then a reasonable proportion of them will probably choose to do so. In fact we've had this in the UK for decades - you can subscribe to an "Economy 7" tariff, which gives you more expensive than normal power during the day and then 7 hours of cheap power each night. Unfortunately the "more expensive than normal during the day" bit tends to make it an unrealistic tariff for anyone who doesn't use electric storage heaters.
I can, however, see a possibility for automated algorithms deciding when to use power - e.g. telling the dishwasher "automatically do the washing up when it'll be cheapest" and having it sit there monitoring the instantaneous electricity prices and automatically doing the right thing. Or loading the washing machine with instructions like "this washing needs to be done some time in the next 3 days, do it when the power is cheapest". This is essentially the same as having computers doing stock-market trading. The interesting bit will be when many people have the same device and they all decide the power is cheapest at the same time, causing a surge in demand and raising the prices.
is more important than the right not to be offended.
Depends on the intent - if you say something and someone gets offended then fair enough; if you say something with the _intent_ for it to offend someone then that is not cool.
I'm conflicted. On the one hand my initial response was like yours. Yet on the other I don't see why, if you were trying to stop a serious threat, spies shouldn't be able to monitor these communications in principle, with some clear restrictions:
Firstly we have the perennial problem that the security services are allowed to spy on anyone with very little oversight. If they want to spy on someone they should be required to get a court order, and that court order should be made public so that everyone can see what they are doing. If the court order cannot be immediately made public for legitimate security reasons then it should be made public as soon as possible (i.e. certainly within a year, preferably sooner). Furthermore, information gathering should not start until that court order is issued - i.e. there should be no requirement for ISPs/telcos to log and retain traffic "just in case" it is needed at a later date.
So given that we already have this problem, further extending the powers of the seucrity services seems like a bad plan.
Futhermore, this stuff is always justified as "to stop a serious threat", and yet there seems to be very little evidence that there are lots of "serious threats" that need stopping. And as always, this stuff is always spun as "to stop the criminals" and attention is diverted from the fact that not everyone who uses a lawyer is a criminal.
1/ If the information gathered by spying was specifically barred from being used in court
Even if you can't use the evidence in court, it can be used to influence a court case, either by directing a line of questioning, or helping with parallel construction of evidence.
2/ If additional authority had to be granted by the judiciary for the act
3/ If there were clear checks and balances in place to deal with abuse.
Except these things clearly aren't happening, or even intended to happen.
The whole point of communications with your lawyer being privalidged is that you can have a completely frank discussion with them in order to prepare your defense. This cannot happen if you are constantly having to avoid incriminating yourself - one of the reasons for getting a lawyer is that they can tell you when to stop talking to avoid that, so if you can't discuss this with them then that seriously harms your defense. If the authorities believe that there is no merit in allowing private legal discussions then this should be true on both sides - the prosecution should be required to make all their discussions public too. As it stands, the laws are very one-sided and stack the deck against anyone the authorities decide to attack, guilty or not.
It seems impractical because the transaction takes about a second at best
Not true - I can't find the link at the moment, but the London Underground has been working with card issuers for a few years to ensure the cards are quick enough to be used to pay for journies during rush hour. ISTR they required transactions to complete in under about 300ms.
so someone would have to shove up against you and hold their reader against your pocket for the full second to make it work.
Not uncommon in a crowded place. The article suggested performing the attack at an airport since foreign currency transactions would not be unusual - if you've ever waited in line while going through airport security you'd realise that an attacker would have ample opportunity to stand right next to you for many seconds if not minutes.
That is assuming you only have one NFC card in your wallet, otherwise interference as multiple cards try to respond will scupper the attack anyway.
Untrue. The protocol allows the card reader to enumerate multiple cards at the same time, select which to talk to and to freely switch between them. Multiple cards are not an issue here.
Really? Considering normal moon missions need a significant boost to get to the moon, how did a commercial satellite do that?
Inclination changes are really expensive. By comparison, a Hohmann transfer orbit to the moon and back can be cheaper and can use the moon's gravity to change inclination. The Apollo missions used a free-return trajectory rather than Hohmann transfer orbit since they needed to get to the moon quickly (don't want a bunch of astronauts spending a few months in deep space), which is why they needed significantly more delta-V. Its worth noting that a commercial satellite generally has a significant amount of station keeping fuel since its expected to stay in service for many years - sacrificing a few years of service is reasonable if the alternative is to completely abandon the satellite because its in the wrong orbit.
I think in the incident in question, someone (Lockheed?) ended up patenting the manouver...
I can't give anyone a non-GPL licence to this work, which is what they were demanding.
IANAL, but are you sure this is the case? I believe that in my country (Norway) at least, you're still the sole proprietor of your IP.
I am the owner of any code I sumbit to the Linux kernel, *but* it is also considered a "derived work" of the rest of the kernel (which means, legally, I'm not the *sole* owner) and therefore the GPL applies.
Did they want to gain exclusive rights to code you'd already published under the GPL?
The contract was non-specific on what code they were talking about - it was a blanket "you will give us a perpetual nonexclusive licence to do what we want with any IP in your ownership which you produced before, after or during your employment with us" (or words to that effect - I can't recall the exact wording).
I don't know how legal it was - as I mentioned, the company in question was already ignoring their TUPE obligations. However, legal or not, I saw no merit in signing it, so I didn't.
Does the GPL preclude that you grant, for instance, a BSD or Apache license for code which you wrote yourself?
The GPL doesn't prevent dual-licensing code for which you are the sole owner (i.e. you wrote it, or the copyright was assinged to you; and it is not derived from anyone else's code). This even extends to commercial licences - i.e. I can write some code and release it under GPL, at the same time as selling a paid-for licence with non-GPL terms to a few people. However, when you contribute code to an existing project, it is usually considered to be a "derived work" since it almost always makes use of existing parts of that project's code - therefore the writer of contributed code would seldom be considered the sole owner, so whatever licence it is released under would need to be fully compatible with the licence used on the rest of the project. This generally precludes dual-licencing code that has been contributed to a GPLed project.
Much like other copyrighted stuff like music - if you make a song that is derived directly from someone else's song then you can't just blindly release it yourself - generally to release a derived song you need to get a licence to do so from the owner of the original song.
The obvious problem with that is that your past work may very well be someone else's property.
You may simply have no standing to grant a license to your past work.
Yes, one of the reasons I cited for refusing to sign it is that a lot of my past work is stuff like Linux kernel coding (which automatically inherits the GPL) - I can't give anyone a non-GPL licence to this work, which is what they were demanding.
But aside from that - if someone wants a licence to all the work I do over the entire course of my life outside of my employment with them, they can damned well pay me a salary for my entire life too! As far as I'm concerned, an employer is entitled to any work I do during my contracted working hours (usually 37.5 hours a week - 09:00 - 17:30, excluding lunch hour); if they want to claim ownership on anything I do in the other 130.5 hours a week then they are going to need to pay me 3.5 times as much for the same hourly rate.
FWIW, this was during a contract renegotiation after my department had been sold off - in theory the new owner needed to comply with TUPE legislation but they had issued a "sign the new contract or be fired" order (which is illegal). However, when I refused to sign, they did reword the contract to remove that clause, so I guess they were trying to do *something* to avoid getting sued.
It's worse than that. If the company you apply for a job at has any interest in the patents, chances are that they will not offer you a job. The problem is that you selling/licensing patents to them while an employee will easily be seen as a conflict of interest.
If they want you and the patents, I believe they may require you to sign over any and all IP to them as terms of employment, compensated by a signing bonus.
I've not got any patents, but at one point I was handed an employment contract that demanded I grant a licence to all my past and future work (which I refused to sign), so you could very well be right. (I'm in the UK, although the company in question was headofficed in Canada)
It does not add measurable value to your home to be fully wired. It does help a house stay on the market a shorter amount of time.
I'm not convinced enough people care about a wired house for it to make a measurable difference on anything at all. In fact, I'd go so far as to say that the vast majority of people don't understand why they might want wired rather than wireless (and an increasing proportion of the population are using computing devices that can't even be connected to a wired network these days too).
As a techy, whether or not a house is wired is probably something that I would care about, but there are an awful lot of other far more important things I would be interested in too, to the point that I doubt that the wiring is likely to make much of a dent in my decision making process.
More likely, I suspect, is someone non-technical looking at my house and concluding that the neatly installed comms cabinet containing the patch panel, VDSL modem, etc. is an unsightly waste of space and that they need to do work to remove it.
I've seen what normal electrical subs do with cat5. It's not pretty. It's downright scary. It's the stuff of nightmares. Assuming that you can just use cat5 used for phone runs may not necessarily work out for you.
This is why I had my house wired with structured cabling. That kind of dense cable bundle can't quite be abused as readily as a single strand of cat5.
I'm not sure I would trust a normal sparky to do networking stuff... Friend of mine just got an office rewired - sparky was told to put in a phone socket and 2 ethernet sockets per desk. Cat6 was specified throughout, but it sounds like the bosses (who were telling the sparky what to do) never actually gave him that specification. Ended up with cat3 for the phones and cat6 for ethernet.
Even given that the sparky seemingly hadn't been explicitly told to use cat6 for the phones, who the hell cables up a modern office in cat3 these days?! Half the Ethernet sockets were punched down with the pairs in the wrong order, and next to no slack cable was left at the patch panel end (which meant they all had to be extended through a punch block).
Some of the blame can certainly be placed on my friend for not staying on top of exactly what the sparky was doing, but at the end of the day the sparky obviously had no clue. Nope, get a real network cable installer to do this stuff!
If you're doing it yourself, you can probably do the job equally well as I can, but you'll need to go down to home depot or lowes and get the cheapest RJ45 crimper you can find (about $20) maybe a 300 foot spool of cat5e wire (as cheap as $20) a box of RJ45 terminators (about $20) modular jacks (about $5 each) and modular faceplates (about $1 each.)
Go spend about an hour on youtube to see how to crimp RJ45 ends (it's actually easier than it sounds) and stick with the 568-b standard for all ends. Don't worry about crossover, straight through, etc. Every time I hear people try to be "smart" and talk about doing it "right" I kind of chuckle, and here's why: Part of the gigabit ethernet standard (that is, to receive IEEE 802.3 certification for gigabit) the switches AND the ethernet ports MUST provide the auto-MDIX feature, so fretting about crossover is pointless.
105 metres of cable doesn't sound like a lot for a 5 bedroom house to me. I have part-wired my 2 bedroom house (finishing the job is on my todo list) and 105 metres of cat6 got me 4 sockets in the living room and 3 sockets in the office, all running back to a cabinet in the office. At some point I will finish the job to a total of 4 sockets in the living room, 2 in each bedroom, 2 in the kitchen, 2 in the stair well (for one of the wifi APs), 2 in the attic, 6 - 10 in the office.
As for crossing over - IMHO it's important to do straight through everywhere, because you may not always be running Ethernet over the structured cabling. My POTS/VDSL is terminated in the living room, but I don't want equipment there so that gets patched straight into the structured cabling and the VDSL modem and PBX is in the data cabinet. Similarly I have a POTS handset plugged into the structured cabling (which the PBX automatically bridges directly to the POTS line if there's a power outage).
The biggest pain is running hidden cables - running them under the upstairs floors involves removing furniture and pulling up carpets and floorboards, and getting the cable from the floor space into the stud walls involves drilling a hole with a 90 degree bend in it because the floor cavity of each room has a support beam at the edge, and the stud walls are capped with timber, so any cables need to go through the support beam, make a 90 degree bend, then go through the cap. I'm sure it was fine when the power cables were put in during construction since it would've been done before the plasterboard was put up, but trying to retrofit is a problem.
Secondly, my living room has a load of sound insulation bonded to the back of the plasterboard on one of the walls, filling the cavity - I've pretty much written off any idea of putting sockets in that wall since it would involve chasing the cable into the wall rather than just running it through the cavity. I might reconsider when it's time to decorate, but I'm not about to trash the existing decor. There are a couple of places where I've just ended up putting a section of surface conduit on the wall because there was just no sane way to run hidden cables without doing serious damage to the decor.
You know, AC has a point there. It seems that every slightly larger framework coming to Linux gets opposed. To me the funniest part is that many of the opponents do not even seem to precisely know why they are opposing the thing, they just quickly learn to robotically chant the same thing than everyone else.
I think Pulse Audio got a bad reputation because it was pushed on people way too early. I can certainly remember upgrading a few systems and finding my audio completely broken in a practically unfixable way (short of wiping and downgrading again) because distros had rolled out PulseAudio and it was so well integrated into stuff that you couldn't just rip it out again. These days it seems to work well and more or less sets out to do what it was designed to do (although I don't think I get a huge amount of benefit from PulseAudio over plain ALSA in day to day use).
I've only used Windows 8 once, but for me it fell down on the "discoverability" criteria - in a GUI, things need to be easily discoverable without googling or consulting the manual - at least, the simple stuff does. First time I used Windows 8 I got presented with the start screen, clicked on the IE logo and up popped the Metro version of IE.... Now what? There's no "Start" menu or anything especially obvious to get you back to the start screen. I pressed ctrl+alt+del and was pleasantly surprised to find that after I killed off IE I was left at a familiar classic Windows desktop. GUIs shouldn't ever leave you in a situation where it isn't obvious how to get back to wherever you came from, and IMHO Metro failed on that count - sure I would've figured it out and got used to it through constant use, but the initial impression is bad, and first impressions count.
Stuff like systemd and udev probably get some backlash because they are quite complex, and are replacements for very simple systems so there is a really steep learning curve that practically never existed before. You get a lot of "I just need to do $trivial_thing, it would've been easy under $old_system but now I'm having to spend forever reading the manual for $new_system!" I wouldn't stand by the "people hate change" argument, more "people hate change when things don't work". Nothing worse than wasting 2 hours trying to fix a problem with the new system that would've been sorted in 5 minutes if you were using something more familiar.
Gnome's reduction of customizability began in the early millennium when it partnered with some large companies who had carried out formal UI studies and found that for the vast majority of users, options only confuse them.
And it's probably true - give most people a system that is set up for them and they are probably happier than having lots of options. The problem with this, of course, is that "set up for them" is different for each user, and out of the box it isn't really set up right for anyone.
They also made some bonkers design decisions that didn't reduce the configurability but not the complexity of the UI - for example, for a long time they claimed no one needed to turn off DPMS, so the "turn off screen" option just had a list of timeouts (5 minutes, 30 minutes, etc). Sticking a "Never" option in there wouldn't have increased the complexity at all because thats exactly where you would expect to find that option.
And bonkers design decisions that increase the complexity of the UI for no reason - for example, how do you suspend a machine? Oh that's right you press alt while the system menu is open and the power-off button changes to a suspend button. That's a completely non-discoverable design - the only way you're going to figure that out is by reading the manual. How is it better than just sticking a "suspend" button in the menu too, or adding suspend options to the power-off dialogue?
I still haven't figured out how to properly control the screen brightness when on battery power - as of a few versions ago, my laptop screen automagically goes dim when I unplug power. I can turn it up again, but the next time I unplug it it goes dim again. There's nothing in the power or display settings to configure this - the "power" config page doesn't specify separate settings for mains or battery mode; there's just one slider for screen brightness which doesn't seem to control the default brightness for when you're on battery.
And how do you get stuff like you IM client to start automagically when you log in? When Gnome 3 first appeared, it had empathy built into the UI so it was running all the time. These days I have to manually start it up when I log in - can't see any "auto start" button anywhere.
I wouldn't consider things like disabling the DPMS, suspending a laptop, telling the screen not to go unreadably dim every time you unplug the power, or wanting to auto-start background stuff like your IM client on login, to be "power user" tweaks.
The interesting trend is that it seems to take losing users/slow adoption in droves and mass rioting to get the ball rolling.
Both gnome 3 and windows 8 have seen their user bases outright revolt over their UI changes, and both largely ignored it as "people hate change but they'll learn to love it" until numbers started actually dropping significantly and people started leaving.
It seems to be really good PR actually... Everyone says "Windows 10 is really good", and quietly ignoring the "...because they ripped out all the crap Windows 8 introduced, leaving it identical to Windows 7" bit.:)
To be honest, I don't really buy the "people hate change" thing - sure, some people hate change, but a lot of the time changes are good. Change for the sake of change is often bad, but a lot of change doesn't fit into that category and actually improves things. From my perspective, I think Gnome 3's UI is pretty good - I really like the fundamental design. What I dislike about Gnome 3 isn't the basic design, its that they seem to think that making everyone use dconf is more "user friendly" than providing a proper configuration UI that actually lets you.. uhm.. configure it.
Slashdot Mirror
Google should do whatever it wants. After all, if I get annoyed enough by Google Chrome, I'll just switch back to Firefox or Opera. Only the ChromeOS/ChromeBook/ChromeBox users may be screwed (because they've made the mistake of locking their hardware to a specific vendor browser).
IE taught us that this kind of thing doesn't happen quickly - web developers _still_ have to deal with IE's buggy rendering, despite good alternatives having been available for 15 years. Ok, IE has got better but it's still not great. Users don't see this stuff as a browser problem - if your website doesn't work right then the users see it as a problem with your website.
StartSSL.com gives free Class1 and is preinstalled in every modern browser
And whilst I use StartSSL, it's a pain that you can't get free wildcard certs for your domain...
Well, it is pretty much like real life, but I'm not sure we want to be teaching kids "this is the crap you can expect fom life" rather than inspiring them to do more.
I did take a slight exception to this though:
But Steven and Brian are also everything frustrating about the tech industry. Steven and Brian represent the tech industry assumption that only men make meaningful contributions.
As far as I can tell from the story, Steven and Brian did nothing wrong at all - clueless Barbie fucks things up and then asks them to fix it, which they do. This bit of the story would probably be pretty similar if you replace Barbie with any clueless person (male or female) who's just infected a bunch of computers with a virus. What were they supposed to do in this situation?
An elite crowd trying to force on everyone else what they think is the right way? Thats one of the many reasons people are against systemd!
The maintainers (you call them "an elite crowd") of some distros have made the decision to use systemd because they think that's the right thing to do - someone has to make the decision, and if not the maintainers, who? Or would you prefer that the maintainers decide to do something that they think isn't right?
No one is forcing anyone to use systemd - the source is there for anyone to use as they see fit; Some distros have decided that systemd is the right way to go, some have decided to use other inits, you can either choose the distro (from a wide selection) that suits your purposes the most, or you can even make your own, no one is forcing you to use one particular distro.
Note: I don't really have any opinions about systemd, I currently use Fedora and it seems to work ok, but if I have problems then I can switch distros.
One thing I don't understand is how in the hell it is considered ok to have this in Debian STABLE? Maybe, in Fedora or OpenSuse but Debian stable???!
Why not Debian Stable? Red Hat Enterprise Linux uses systemd, so it must be good enough for enterprise use, so why it it not good enough for Debian Stable?
"systemd does the right thing by stopping normal boot and just boot into a safe, minimal shell. A quick glace in the log file (journal) will instantly tell you (using red letters for emphasis) that fstab is broken in such and such a way. A quick edit with Vim can then solve the problem." - did you miss these lines in his comment? Just how "far" is "far enough" ?
Well that would depend... If its your desktop machine then popping a shell on the screen would probably work(*). If it's a headless networked device then you're going to need the NICs brought up and sshd started.
(*) This isn't especially user friendly though... how about firing X up and having a nice GUI thing to fix the problem?
This kind of tight coupling is unheard of in Linux history.
Not true at all - stuff has been tightly coupled plenty of times in the past. Lots of stuff is very tightly coupled with udev these days, for example. And whilst I will agree that tight coupling is bad, its sometimes hard to see how it could be avoided.
The use of System V init allowed Linux to be comfortablef for UNIX admins looking for a less expensive or more widely installable solution, and the end of the use of System V init means that Linux is starting to head away from the UNIX operating systems.
Linux has been heading away from Unix systems for a long time. As a long-time Linux user, on the odd occasion that I have to deal with the likes of Solaris I find it feels *very* backwards by comparison... It's almost like going back to the 1980s...
What if it was someone attacking your sshd and making it crash when it failed?
By automatically restarting it, you just allow the attacker to continue trying to exploit it.
By automatically restarting it, you don't solve the issue that makes it crashing.
By automatically restarting it, you, most of the time, don't even see it restarted, so really not giving you any way to solve the real problem.
It's not that I don't find process monitoring interesting, it's just that automatically restarting can bring more problems than it solves.
As with any service, the "correct" action upon a crash is probably dependent on what the machine is actually supposed to be doing. Take for example, a dedicated web server - having Apache do down when under attack and not attempt to recover would be bad since the attacker will have successfully caused a denial of service with very little effort. Compare to a private telephone exchange, for example, which is running a web server purely for management purposes - a crashed web server is not a disaster, the whole thing keeps doing its primary job without it, so automatically restarting the crashed web service _may_ not be the best plan.
So I guess the answer here is "it depends" and therefore the administrator should be able to choose either option, so selecting an init system that doesn't support one of the options would be bad.
In the case of sshd, since it is potentially the only way to safely fix a broken server, allowing it to die permanently seems like a bad option to me. A better option would probably be to restart it and firewall off all but a few "safe" IP addresses. That way the administrator can still access the server from one of those IPs and the attacker can't cause any more damage.
A well behaving daemon shouldn't be restarted (except maybe for rereading config files), it should start and stay that way. If it crashes randomly, then you might try to find the bug.
Whilst I agree that you should fix a crashy service rather than restarting it each time it breaks, there are nver the less reasons why you may want to auto-restart the service:
- In the real world, you can't just shut down a service until a bug has been fixed; you need to continue running it as best you can while the problem is being looked into and fixed. So a stop-gap measure may be necessary.
- Whilst you may believe some software to be bug-free, this may not be the case, and in some cases it would be disasterous to discover that thre is a bug by finding a service permanently go down. Far better to restart it and log the error.
- Bits _do_ occasionally get flipped in memory or registers, so software may well occasionally crash through no fault of its own. It is reasonable to have something in place to mitigate this should it ever happen.
So yes, I agree, if a service is crashing all the time then it needs to be fixed, but that doesn't mean that you should abandon all possibility of recovering from an unexpected crash.
Comcast offered a $50 cash card if we signed up for internet service with them. We signed up in May, and the card never came. We called and they denied that they ever offered the card. A few more calls later, they agreed that they offered it and said they would send it. It never came. Last month, five months, a final call was made and the card arrived.
Clearly they have a strategy of screwing customers, either through intentional scripting or extreme negligence.
Not just telecomms companies - I'm currently being screwed over by Npower. I was a customer for 18 months, during that time I got a single correct bill and had to spend hours chasing them to get the others corrected. My original contract gave me a discount on the billing anniversary (January) - they never credited the discount, so I raised a dispute in January and they agreed to credit it. Except it never got credited. In the summer I left them as a customer, filed a complaint (about the shiteness of their service and about the unpaid discount) and refused to pay the final bill as it was incorrect. They responded to my dispute, replying to my complaint about the poor service but ignoring my points about the incorrect bill. I re-raised the complaint and got an automated "we'll respond in under 10 days" reply - never got an actual response so I followed it up 10 days later, again no response. The billing department are now threatening to take out a court summons against me for the unpaid (incorrect) bill. The billing department say they can't do anything about the disputed bill or the threatened court action and that I will need to raise a complaint with the complaints department, completely ignoring the fact that I've already tried to raise a complaint several times and the complaints department won't respond.
So now I've sent 50MB of paperwork and telephone recordings to the regulator in the hope that they can beat some sense into Npower before I have to waste yet more time defending myself in court.
I'm left wondering if this is actually incompetence, or if their corporate policy is to conveniently "forget" to pay discounts that were promised a year ago in the hope that most people won't notice.
It's somewhat like buying a last minute airline ticket. If people were unwilling to pay more for a last minute ticket, all tickets would cost more (fine) but it would be impossible (because the airlines would price tickets to insure every seat was sold - or oversold - many hours before wheels up to minimize the risk of a single empty seat) to get a ticket on a commercial airliner to get to mom's bedside 1500 miles away before she expires.
Airline tickets are an interesting one; and other products that have a relatively inflexible supply - i.e. the costs of flying the plane are basically the same whether or not that seat is sold, and similarly you can't sell more seats than you have, so the supply is inflexible. There are two opposing forces at work here:
1. The airline wants to ensure that every seat is sold, since filling a seat at the last minute increases profit, even if it is sold for below cost (the plane is flying anyway, the costs can't be avoided, selling the seat rather than flying with it empty is beneficial no matter how cheaply you sell it). This is going to tend to push the prices down for "last minute" sales as the airline tries to attract sales.
2. The passengers that need to get somewhere at short notice are willing to pay a premium. This is going to push the "last minute" prices up as the airlines cash in on this willingness to pay over the odds.
Figuring out which of these forces wins is certainly a non-trivial exercise.
By increasing prices when demand approaches the absolute maximum supply, consumers will reduce demand quickly (good, since supply can't be increased quickly). When power gets expensive enough, they will shut off rooms, wear more sweaters, turn lights off, instead of cooking a fancy dinner they will nuke something in the microwave and use disposable utensils (or, just wait to wash them until the next day), they will sit around in a single room and talk instead of playing on their computer or watching TV in individual rooms. Demand is extremely elastic, supply is inelastic at the top end. In extreme cases, they will shutdown their entire house (using winter shutdown procedures as needed) and gather in friends and neighbor's houses (perhaps, splitting the cost of the very expensive power during those times).
I think expecting people to monitor electricity prices on a minute by minute basis and change what they are doing _now_ is (largely) not realistic - virtually no one is going to look at the electricity price before deciding to put the TV on, for example. What is realistic is getting people into a routine - if people know that it's always cheaper for them to put the dish washer / washing machine / whatever on over night, then a reasonable proportion of them will probably choose to do so. In fact we've had this in the UK for decades - you can subscribe to an "Economy 7" tariff, which gives you more expensive than normal power during the day and then 7 hours of cheap power each night. Unfortunately the "more expensive than normal during the day" bit tends to make it an unrealistic tariff for anyone who doesn't use electric storage heaters.
I can, however, see a possibility for automated algorithms deciding when to use power - e.g. telling the dishwasher "automatically do the washing up when it'll be cheapest" and having it sit there monitoring the instantaneous electricity prices and automatically doing the right thing. Or loading the washing machine with instructions like "this washing needs to be done some time in the next 3 days, do it when the power is cheapest". This is essentially the same as having computers doing stock-market trading. The interesting bit will be when many people have the same device and they all decide the power is cheapest at the same time, causing a surge in demand and raising the prices.
is more important than the right not to be offended.
Depends on the intent - if you say something and someone gets offended then fair enough; if you say something with the _intent_ for it to offend someone then that is not cool.
I'm conflicted. On the one hand my initial response was like yours. Yet on the other I don't see why, if you were trying to stop a serious threat, spies shouldn't be able to monitor these communications in principle, with some clear restrictions:
Firstly we have the perennial problem that the security services are allowed to spy on anyone with very little oversight. If they want to spy on someone they should be required to get a court order, and that court order should be made public so that everyone can see what they are doing. If the court order cannot be immediately made public for legitimate security reasons then it should be made public as soon as possible (i.e. certainly within a year, preferably sooner). Furthermore, information gathering should not start until that court order is issued - i.e. there should be no requirement for ISPs/telcos to log and retain traffic "just in case" it is needed at a later date.
So given that we already have this problem, further extending the powers of the seucrity services seems like a bad plan.
Futhermore, this stuff is always justified as "to stop a serious threat", and yet there seems to be very little evidence that there are lots of "serious threats" that need stopping. And as always, this stuff is always spun as "to stop the criminals" and attention is diverted from the fact that not everyone who uses a lawyer is a criminal.
1/ If the information gathered by spying was specifically barred from being used in court
Even if you can't use the evidence in court, it can be used to influence a court case, either by directing a line of questioning, or helping with parallel construction of evidence.
2/ If additional authority had to be granted by the judiciary for the act
3/ If there were clear checks and balances in place to deal with abuse.
Except these things clearly aren't happening, or even intended to happen.
The whole point of communications with your lawyer being privalidged is that you can have a completely frank discussion with them in order to prepare your defense. This cannot happen if you are constantly having to avoid incriminating yourself - one of the reasons for getting a lawyer is that they can tell you when to stop talking to avoid that, so if you can't discuss this with them then that seriously harms your defense. If the authorities believe that there is no merit in allowing private legal discussions then this should be true on both sides - the prosecution should be required to make all their discussions public too. As it stands, the laws are very one-sided and stack the deck against anyone the authorities decide to attack, guilty or not.
It seems impractical because the transaction takes about a second at best
Not true - I can't find the link at the moment, but the London Underground has been working with card issuers for a few years to ensure the cards are quick enough to be used to pay for journies during rush hour. ISTR they required transactions to complete in under about 300ms.
so someone would have to shove up against you and hold their reader against your pocket for the full second to make it work.
Not uncommon in a crowded place. The article suggested performing the attack at an airport since foreign currency transactions would not be unusual - if you've ever waited in line while going through airport security you'd realise that an attacker would have ample opportunity to stand right next to you for many seconds if not minutes.
That is assuming you only have one NFC card in your wallet, otherwise interference as multiple cards try to respond will scupper the attack anyway.
Untrue. The protocol allows the card reader to enumerate multiple cards at the same time, select which to talk to and to freely switch between them. Multiple cards are not an issue here.
Really? Considering normal moon missions need a significant boost to get to the moon, how did a commercial satellite do that?
Inclination changes are really expensive. By comparison, a Hohmann transfer orbit to the moon and back can be cheaper and can use the moon's gravity to change inclination. The Apollo missions used a free-return trajectory rather than Hohmann transfer orbit since they needed to get to the moon quickly (don't want a bunch of astronauts spending a few months in deep space), which is why they needed significantly more delta-V. Its worth noting that a commercial satellite generally has a significant amount of station keeping fuel since its expected to stay in service for many years - sacrificing a few years of service is reasonable if the alternative is to completely abandon the satellite because its in the wrong orbit.
I think in the incident in question, someone (Lockheed?) ended up patenting the manouver...
I can't give anyone a non-GPL licence to this work, which is what they were demanding.
IANAL, but are you sure this is the case? I believe that in my country (Norway) at least, you're still the sole proprietor of your IP.
I am the owner of any code I sumbit to the Linux kernel, *but* it is also considered a "derived work" of the rest of the kernel (which means, legally, I'm not the *sole* owner) and therefore the GPL applies.
Did they want to gain exclusive rights to code you'd already published under the GPL?
The contract was non-specific on what code they were talking about - it was a blanket "you will give us a perpetual nonexclusive licence to do what we want with any IP in your ownership which you produced before, after or during your employment with us" (or words to that effect - I can't recall the exact wording).
I don't know how legal it was - as I mentioned, the company in question was already ignoring their TUPE obligations. However, legal or not, I saw no merit in signing it, so I didn't.
Does the GPL preclude that you grant, for instance, a BSD or Apache license for code which you wrote yourself?
The GPL doesn't prevent dual-licensing code for which you are the sole owner (i.e. you wrote it, or the copyright was assinged to you; and it is not derived from anyone else's code). This even extends to commercial licences - i.e. I can write some code and release it under GPL, at the same time as selling a paid-for licence with non-GPL terms to a few people. However, when you contribute code to an existing project, it is usually considered to be a "derived work" since it almost always makes use of existing parts of that project's code - therefore the writer of contributed code would seldom be considered the sole owner, so whatever licence it is released under would need to be fully compatible with the licence used on the rest of the project. This generally precludes dual-licencing code that has been contributed to a GPLed project.
Much like other copyrighted stuff like music - if you make a song that is derived directly from someone else's song then you can't just blindly release it yourself - generally to release a derived song you need to get a licence to do so from the owner of the original song.
The obvious problem with that is that your past work may very well be someone else's property.
You may simply have no standing to grant a license to your past work.
Yes, one of the reasons I cited for refusing to sign it is that a lot of my past work is stuff like Linux kernel coding (which automatically inherits the GPL) - I can't give anyone a non-GPL licence to this work, which is what they were demanding.
But aside from that - if someone wants a licence to all the work I do over the entire course of my life outside of my employment with them, they can damned well pay me a salary for my entire life too! As far as I'm concerned, an employer is entitled to any work I do during my contracted working hours (usually 37.5 hours a week - 09:00 - 17:30, excluding lunch hour); if they want to claim ownership on anything I do in the other 130.5 hours a week then they are going to need to pay me 3.5 times as much for the same hourly rate.
FWIW, this was during a contract renegotiation after my department had been sold off - in theory the new owner needed to comply with TUPE legislation but they had issued a "sign the new contract or be fired" order (which is illegal). However, when I refused to sign, they did reword the contract to remove that clause, so I guess they were trying to do *something* to avoid getting sued.
It's worse than that. If the company you apply for a job at has any interest in the patents, chances are that they will not offer you a job.
The problem is that you selling/licensing patents to them while an employee will easily be seen as a conflict of interest.
If they want you and the patents, I believe they may require you to sign over any and all IP to them as terms of employment, compensated by a signing bonus.
I've not got any patents, but at one point I was handed an employment contract that demanded I grant a licence to all my past and future work (which I refused to sign), so you could very well be right. (I'm in the UK, although the company in question was headofficed in Canada)
It does not add measurable value to your home to be fully wired. It does help a house stay on the market a shorter amount of time.
I'm not convinced enough people care about a wired house for it to make a measurable difference on anything at all. In fact, I'd go so far as to say that the vast majority of people don't understand why they might want wired rather than wireless (and an increasing proportion of the population are using computing devices that can't even be connected to a wired network these days too).
As a techy, whether or not a house is wired is probably something that I would care about, but there are an awful lot of other far more important things I would be interested in too, to the point that I doubt that the wiring is likely to make much of a dent in my decision making process.
More likely, I suspect, is someone non-technical looking at my house and concluding that the neatly installed comms cabinet containing the patch panel, VDSL modem, etc. is an unsightly waste of space and that they need to do work to remove it.
I've seen what normal electrical subs do with cat5. It's not pretty. It's downright scary. It's the stuff of nightmares. Assuming that you can just use cat5 used for phone runs may not necessarily work out for you.
This is why I had my house wired with structured cabling. That kind of dense cable bundle can't quite be abused as readily as a single strand of cat5.
I'm not sure I would trust a normal sparky to do networking stuff... Friend of mine just got an office rewired - sparky was told to put in a phone socket and 2 ethernet sockets per desk. Cat6 was specified throughout, but it sounds like the bosses (who were telling the sparky what to do) never actually gave him that specification. Ended up with cat3 for the phones and cat6 for ethernet.
Even given that the sparky seemingly hadn't been explicitly told to use cat6 for the phones, who the hell cables up a modern office in cat3 these days?! Half the Ethernet sockets were punched down with the pairs in the wrong order, and next to no slack cable was left at the patch panel end (which meant they all had to be extended through a punch block).
Some of the blame can certainly be placed on my friend for not staying on top of exactly what the sparky was doing, but at the end of the day the sparky obviously had no clue. Nope, get a real network cable installer to do this stuff!
If you're doing it yourself, you can probably do the job equally well as I can, but you'll need to go down to home depot or lowes and get the cheapest RJ45 crimper you can find (about $20) maybe a 300 foot spool of cat5e wire (as cheap as $20) a box of RJ45 terminators (about $20) modular jacks (about $5 each) and modular faceplates (about $1 each.)
Go spend about an hour on youtube to see how to crimp RJ45 ends (it's actually easier than it sounds) and stick with the 568-b standard for all ends. Don't worry about crossover, straight through, etc. Every time I hear people try to be "smart" and talk about doing it "right" I kind of chuckle, and here's why: Part of the gigabit ethernet standard (that is, to receive IEEE 802.3 certification for gigabit) the switches AND the ethernet ports MUST provide the auto-MDIX feature, so fretting about crossover is pointless.
105 metres of cable doesn't sound like a lot for a 5 bedroom house to me. I have part-wired my 2 bedroom house (finishing the job is on my todo list) and 105 metres of cat6 got me 4 sockets in the living room and 3 sockets in the office, all running back to a cabinet in the office. At some point I will finish the job to a total of 4 sockets in the living room, 2 in each bedroom, 2 in the kitchen, 2 in the stair well (for one of the wifi APs), 2 in the attic, 6 - 10 in the office.
As for crossing over - IMHO it's important to do straight through everywhere, because you may not always be running Ethernet over the structured cabling. My POTS/VDSL is terminated in the living room, but I don't want equipment there so that gets patched straight into the structured cabling and the VDSL modem and PBX is in the data cabinet. Similarly I have a POTS handset plugged into the structured cabling (which the PBX automatically bridges directly to the POTS line if there's a power outage).
The biggest pain is running hidden cables - running them under the upstairs floors involves removing furniture and pulling up carpets and floorboards, and getting the cable from the floor space into the stud walls involves drilling a hole with a 90 degree bend in it because the floor cavity of each room has a support beam at the edge, and the stud walls are capped with timber, so any cables need to go through the support beam, make a 90 degree bend, then go through the cap. I'm sure it was fine when the power cables were put in during construction since it would've been done before the plasterboard was put up, but trying to retrofit is a problem.
Secondly, my living room has a load of sound insulation bonded to the back of the plasterboard on one of the walls, filling the cavity - I've pretty much written off any idea of putting sockets in that wall since it would involve chasing the cable into the wall rather than just running it through the cavity. I might reconsider when it's time to decorate, but I'm not about to trash the existing decor. There are a couple of places where I've just ended up putting a section of surface conduit on the wall because there was just no sane way to run hidden cables without doing serious damage to the decor.
If chimps are people, will they be able to vote? Hold political office? Cue the jokes.
Also, does that mean we have to jail them when they commit crimes against each other? (E.g. stealing each other's food, etc).
Windows 8, bad.
PulseAudio, bad.
Wayland, bad...
You know, AC has a point there. It seems that every slightly larger framework coming to Linux gets opposed. To me the funniest part is that many of the opponents do not even seem to precisely know why they are opposing the thing, they just quickly learn to robotically chant the same thing than everyone else.
I think Pulse Audio got a bad reputation because it was pushed on people way too early. I can certainly remember upgrading a few systems and finding my audio completely broken in a practically unfixable way (short of wiping and downgrading again) because distros had rolled out PulseAudio and it was so well integrated into stuff that you couldn't just rip it out again. These days it seems to work well and more or less sets out to do what it was designed to do (although I don't think I get a huge amount of benefit from PulseAudio over plain ALSA in day to day use).
I've only used Windows 8 once, but for me it fell down on the "discoverability" criteria - in a GUI, things need to be easily discoverable without googling or consulting the manual - at least, the simple stuff does. First time I used Windows 8 I got presented with the start screen, clicked on the IE logo and up popped the Metro version of IE.... Now what? There's no "Start" menu or anything especially obvious to get you back to the start screen. I pressed ctrl+alt+del and was pleasantly surprised to find that after I killed off IE I was left at a familiar classic Windows desktop. GUIs shouldn't ever leave you in a situation where it isn't obvious how to get back to wherever you came from, and IMHO Metro failed on that count - sure I would've figured it out and got used to it through constant use, but the initial impression is bad, and first impressions count.
Stuff like systemd and udev probably get some backlash because they are quite complex, and are replacements for very simple systems so there is a really steep learning curve that practically never existed before. You get a lot of "I just need to do $trivial_thing, it would've been easy under $old_system but now I'm having to spend forever reading the manual for $new_system!" I wouldn't stand by the "people hate change" argument, more "people hate change when things don't work". Nothing worse than wasting 2 hours trying to fix a problem with the new system that would've been sorted in 5 minutes if you were using something more familiar.
Gnome's reduction of customizability began in the early millennium when it partnered with some large companies who had carried out formal UI studies and found that for the vast majority of users, options only confuse them.
And it's probably true - give most people a system that is set up for them and they are probably happier than having lots of options. The problem with this, of course, is that "set up for them" is different for each user, and out of the box it isn't really set up right for anyone.
They also made some bonkers design decisions that didn't reduce the configurability but not the complexity of the UI - for example, for a long time they claimed no one needed to turn off DPMS, so the "turn off screen" option just had a list of timeouts (5 minutes, 30 minutes, etc). Sticking a "Never" option in there wouldn't have increased the complexity at all because thats exactly where you would expect to find that option.
And bonkers design decisions that increase the complexity of the UI for no reason - for example, how do you suspend a machine? Oh that's right you press alt while the system menu is open and the power-off button changes to a suspend button. That's a completely non-discoverable design - the only way you're going to figure that out is by reading the manual. How is it better than just sticking a "suspend" button in the menu too, or adding suspend options to the power-off dialogue?
I still haven't figured out how to properly control the screen brightness when on battery power - as of a few versions ago, my laptop screen automagically goes dim when I unplug power. I can turn it up again, but the next time I unplug it it goes dim again. There's nothing in the power or display settings to configure this - the "power" config page doesn't specify separate settings for mains or battery mode; there's just one slider for screen brightness which doesn't seem to control the default brightness for when you're on battery.
And how do you get stuff like you IM client to start automagically when you log in? When Gnome 3 first appeared, it had empathy built into the UI so it was running all the time. These days I have to manually start it up when I log in - can't see any "auto start" button anywhere.
I wouldn't consider things like disabling the DPMS, suspending a laptop, telling the screen not to go unreadably dim every time you unplug the power, or wanting to auto-start background stuff like your IM client on login, to be "power user" tweaks.
The interesting trend is that it seems to take losing users/slow adoption in droves and mass rioting to get the ball rolling.
Both gnome 3 and windows 8 have seen their user bases outright revolt over their UI changes, and both largely ignored it as "people hate change but they'll learn to love it" until numbers started actually dropping significantly and people started leaving.
It seems to be really good PR actually... Everyone says "Windows 10 is really good", and quietly ignoring the "...because they ripped out all the crap Windows 8 introduced, leaving it identical to Windows 7" bit. :)
To be honest, I don't really buy the "people hate change" thing - sure, some people hate change, but a lot of the time changes are good. Change for the sake of change is often bad, but a lot of change doesn't fit into that category and actually improves things. From my perspective, I think Gnome 3's UI is pretty good - I really like the fundamental design. What I dislike about Gnome 3 isn't the basic design, its that they seem to think that making everyone use dconf is more "user friendly" than providing a proper configuration UI that actually lets you.. uhm.. configure it.
There's lots of useful free stuff for people who want to emulate ancient computers at pdp11.org.
Yeah, but that's not in the cloud, and if you're not doing it in the cloud you might accidentally get too much reliability.