Ok.. It depends.... Password rules, like anything, when used within reason CAN increase security.
There has been some research which arrive at the conclusion that yes, indeed, password rules are actually bullshit for security.
As mentioned in the summary, enforcing password rules will actually block provably safe passwords : - a base32 encoded 128bit pure random number. It's mathematically provable to be secure (if done by a cryptography-grade true random number generated, it's a 2^128 security, which is pretty good enough). But it's a 25 character long string of alaphanumeric. So it's not mixed case, and doesn't contain punctuation so it will be rejected by most stupid rules (also some rules have size specified as a range [9 to 16 characters], not a minimum [more than 8]. This will also reject a 25-long password).
As shown in presentations at numerous presentation in conferences such as CCC : - even a complex rule set (Mixed case, must contain numbers and punctiation, at least 9 characters long) will usually give results such as "Denver17!" Which are a lot less secure because they follow a general pattern (The first letter is the single capitalized, number come at the end, punctuation is the last and 9 out 10 times it's a '!' ). Most of these "rule abiding password" follow one of very few such patterns, and patterns are alarmingly easy to crack.
As such, no matter what, rules are a bad idea.
On the other hand, password managers with a generation function (like the above 128-bits equivalent password) are definitely a good idea.
WebEx is a horrible pile of java that isn't very portable (relies on binary JNI libraries for audio/video) and thus doen't work reliably on Linux 64 bits. Adobe Connect is an aweful Flash monster (requieres freshplayer plugin layer under Linux) that crashes at the slightest disturbance (whenever the network devices or the mic sources change).
Meanwhile, all the various "meet"-type services (hangout meet, jitsi meet, etc.) rely on pretty standard HTML5/WebRTC. It works on anything under the sun that is compliant.
(Skype is in a middle ground. The older app might get problems when trying to run. And web Skype relies on ORTC, microsoft's own nih reinterpretation of the standards. So Microsoft Edge-only for now, but efforts are underway to bring it to linux).
It's also a good way to draw attention to yourself... very few people actually use it, so a search of just them is much more practical.
Depends on the region. Europe in general is much more privacy conscious. Specially some central / northern parts (E.g.: Germany, Switzerland). Here you find people who use Tor regularily, just for the sake of keeping their life private.
You can thank these kind of people to help fill tor with background noise traffic. Their are the one you make your reading of the anarchist cookbook less obvious And who make "rounding up all tor users" completely impractical.
(I might not be a pilote but I have a vague notion of the physics involved)
Just like helicopters become autogyros when their engine fails.
(also: as long as the engine didn't take the commercial airplane's wing with it when exploding) (also: and some fighter jets have enough body lift [and avionics to compensate] to afford losing a wing)
I wasn't commenting on the merits of whole plane chute, just mentioning that these are being developed. (But currently, the only reports I've came across are from the manufacturer themselves. So of course these are going to be heavily biased toward "Our products are incredibly useful and save tons of lifes !")
Pay attention that the current default behavious of VBox scripts might open a different kind of vulnerability :
USB-pass-though requires that the VBox process has access to the raw USB device. This is done by the script "/usr/lib/virtualbox/VBoxCreateUSBNode.sh" it creates the appropriate entries in "/dev/vboxusb/" granting them full group access for "vboxusers" Currently this script is called by default by "/etc/udev/rules.d/90-vbox-usb.rules" for any plugged-in device.
That means the raw USB device of *any* USB gizmo is available for *ANY* process that runs with VBox's group credential. This opens quite a big hole. (Virtual Box it self then use a white list so *NOT ALL* device will be available to the Windows 10 VM, only those that you grant access to. But it means potentially any USB device could be hacked by any process running with "vboxusers" group privileges).
A better way is to comment out the insersion rules, and only create the devices for device that you want to be visible to virtual box. Example of a configuration file that only grants access to a few of my devices that rely on windows-only software for firmware updates:
# TomTom Live ATTRS{idVendor}=="1390",ATTRS{idProduct}=="5454", ACTION=="add", RUN+="/usr/lib/udev/VBoxCreateUSBNode.sh $major $minor $attr{bDeviceClass} vboxusers"
As suggested by other discussion threads here around :
You can also achieve the same virtually : "virtually move" the image to an area without any signal.
I.e.: Windows 10 goes into a VirtualBox VM. VM has no network. VM has only CD-ROM (so can read from.iso files you mount) and shared folder (VirtualBox sharing doesn't go through network, so it's not opening windows 10 to remote access, at least not without a collaborating host OS).
You can pass the files and necessary application through shared folders and.ISO respectively.
I would suggest avoiding USB pass-through because it's complex to configure it in a secure way (defaults VBox scripts just make all device visible to any application running with VBox's goup credentials).
You achieve the same as moving a Windows 10 laptop to an area without any signal. (i.e.: No network for Windows 10, no matter what)
But you still get to have an actual connection on your host OS (say a well secured and well kept Linux host).
ICE cars, and their engines, are made from recycled and recyclable materials. {...} And again, Lithium Ion batteries are nothing but toxic landfill when they've outlived their ability to hold a meaningful charge.
And again, over the whole lifetime of the vehicle from manufacturing, through using, to decommissioning, the numbers in current report tend to show that even when factoring everything in, electric cars are still a bit (US) or some (EU) better. (Unless your country is China. Then it's basically the same).
(Don't forget that over their lifetime, ICE will release a significant of poluant too. It's a slow trickle, compared to decommissioning a car battery. But it builds up to very large total quantity. You just spread it more geographically and across time. Which might make it even more difficult to process).
And again, cars don't really make up a huge part of greenhouse gases.
Sorry, no. *THEY DO*. (I'm nitpicking at your usage of the word "huge" vs some superlative like "top contributor").
At 26% (EPA numbers) they are not the *biggest emitter* (that crown goes to power generation according to the same source). But they are still a significantly big chunk of it. Significant enough they deserve to be considered and also addressed.
Our time and effort is better spent elsewhere.
That would have been the case if cars only contributed to say 4% of emissions. That's not the case. Again, even if they are not the *top* emitter, it's still worth spending effort on cars along with all the other emitter that still emits significant chunks of the pollution, even if they are not number 1 position on the ordered list.
Oh, and never mind that on a road trip you're stuck at a charge point for 30 minutes in order to recharge your car enough to drive another 2 hours.
Yeah! Let's drive 8 hours straight without a single break in it! Yay for safety! We human absolutely have no single problem to keep concentrating after hours of driving in a row!
Yup, maybe you're highly organized, doing all your road-trips in teams having at least 3 drivers and driving in shifts between drivers seat, passenger seat and sleeping in the back, while an AI supplement the passenger in their role of helping the driver watching the road.
Or you're one of those guys pretending that he can drive 12hours straight like a machine, only stopping to pee. Simply because up to now they happen by luck not to have had an accident. Even if there are massive study with numbers showing that this is a huge risk, actually.
As battery swapping stations haven't caught yet up, your and your co-driving friends' specific use case isn't well served by vehicles that can only rack up to somewhere around 400-500kms between 30-45min breaks. (And please tell me you're actually of the "taking shifts" variety, not the "Hey, I've driven 12hours only stoping to pee, and I'm still alive, so nobody ever needs to rest !")
You fail to take into account all the other uses cases : - People who actually take 30min break each 2hours because that's the current recommendation to avoid being to distracted/tired/etc. - People who actually drive less than 300km per trip (in fact, most of the trips done with cars tend to be under 100km in Europe. Even the cheapest battery option at my local car-sharing covers it without problems). etc.
Finally, the tech described in the Boeing link isn't a "nice story, bro." It's real and Boeing is actively ramping it up.
But still, there's no gaz station where I can tank my car.
So for now, electric car are still having a better impact on the environment. (Even when you factor everyting including the battery in).
That might change one day. But we still aren't there yet.
and the production method described does exactly what you describe -- use more carbon than it releases.
Couldn't you make a phosphorescent front projection screen?
In theory: yes, specially since there exist phosphors that only react to some laser frequencies (so you could even do RGB). (See Thunderf00t's Youtube video playing with blue phorphor balls and blue lasers). You could in theory spent some resource in researching phosphors that reacts more or less to a laser, like older CRT's phosphor did react to its electron gun.
In practice: it's much cheaper to try to pump-up the frame rate and refresh the vector figure at a rate where the human eye's persistence of vision doesn't notice the flicker (much). (At least that's what I've read people are doing with OpenLase) (And that's how the laser are used in practice in night clubs : drawing the vectors at very high refresh rate).
So now Dark Web means SMTP and FTP? Web? Web. *facepalm*
More seriously : Nope, not at all.
Dark Web has nothing to do with *what* you're talking to (usually websites, so usually HTTP/HTTPS), but *how* you talk to (instead of opening a simple TCP connection between your laptop and a webserver, you send your request using some form of indirect complex hard-to-track routing. e.g.: TOR (The Onion Router)).
On the tor network, there are such things as tor servers which are accessed using a special address. e.g.: Duck Duck Go has one.
From the outside, it doesn't look like your laptop is connecting to DDG's server. It only looks like a bunch of tor nodes passing packets around.
At least in the case of Duck Duck Go, they also have an official DNS name that is registered to them, an IP address that can be backtraced to the server on the rack in the data-center where they are renting (or to the CDN that can then in turn further find them).
But some server only exist as obscure Tor keys. It's very hard to trace them in the real world.
Hence the name "Dark Web". The web of server that is hard to trace cause they don't work over plain vanilla TCP/IP.
It doesn't make sense that 85% of them suddenly stopped being greedy and slimey all of a sudden.
Every criminal thinks they will never get caught, or has already resigned themselves to the idea that it is part of the game. Some might have closed up shop, but 85%? I highly doubt it.
Not 100% of what goes on the dark web is criminal.
Take tor as an example:
Yes, some of them use.onion tor web services for the purpose of hiding ethically-dubious criminal activities. (silk road used to be an example back then).
BUT, some use them for very practical reason like evading censorship (though it is *still* considered illegal in some specific jurisdiction, globally it's not and it's hardly unethical). The popular Duck Duck Go search engine also has a.onion tor server. If you're in China, and want to get informations about Tibet or the Tiananmen Massacre or simply about Taiwan, you can't through a normal search engine. They are either blocked in China or collaborate with the government and only report government-approved informations. On the other hand, if you have access to the tor network (along with VPNs that's one of the few popular solutions to get around China's Great Firewall) you can search the web using Duck Duck Go's onion tor server. (And then read the source using your tor socks proxy).
As needs come and go, or because ressources are limited, some of these non-criminal dark web site are going to disapear.
I wonder if you could just bounce a laser (or simply a highly focused light) off a MEMS mirror or something. Or maybe you'd use multiples?
Such lasers are popular in nightclubs. There are a few hacker who have adapted them to mame (google "openlase")
The main problem is that CRT Vector display have some persistance (the phosphores don't imediately turn black) and the image looks stable, whereas the laser make the image "blink" (unless you find a way to bump up the refresh rate)
My understand is that the NES system was based on timing of scan lines on the CRT
That's one of the two method that it can use. Because the CPU of the NES is very slow, it's reported to be rather shitty on the horizontal axis, it's mostly good only for the vertical axis. (And doesn't work at all with an LCD or a 100Hz CRT. It could at least work for the vertical axis on some laser scanning projectors, but forget about the horizontal axis, all these projectors ping-pong left and right)
Operation Wolf is reported by other to rely on this technique. Most later consoles (basically anything until the Wii) counted exclusively on this method. LCD-/100Hz- workaround used infrared light-sources and image processing (similar to how wii-mote works) and generate appropriate signal to simulate the position. (G1 Light Gun was produced for playstation this way).
The other method consist of quickly blink the screen and the various target black or white. The game knows which target the gun is pointed at by knowing which targets were blinked white at the moment the gun detected a signal. (e.g.: duck hunt turns first the screen black, then each duck is flashed white in turn. When the gun gets a signal it means it's pointing to the currently flashed duck) in theory it should be able to somewhat work with a LCD, provided that all image post-processing is disabled and thus the LCD displays pixels as it is receiving them, so that the gun see exactly what the NES is trying to display. (e.g.: when duck hunt first turn the screen black, the gun must imediately see black - so it know the light source it's pointing to is the screen. If there's a delay, the gun might still see light from the previous frame and decide that the player is trying to cheat by pointing the gun at a lamp)
With current processes for extracting Lithium from the ground resulting in 0.02% lithium and 99.8% dirt that is now contaminated by the toxic chemicals used to extract the lithium and the resource depletion on local water sources as water is shipped to lithium mines in salt flats, plus the fact that electricity to charge the cars likely comes from burning fossil fuels anyway, I'm going to guess the net environmental impact of an electric vehicle over the course of its life is only nominally better than a combustion engine.
According to studies (google is your friend) it all depends on the country.
- In countries that rely very little on greenhouse gazes-emitting power production (e.g.: lots of central and northern european countries) : pollution generated by the manufacture of battery and production of electicity is still a lot smaller than the pollution generated by the production of fuel and burning it in ICE.
- In countries that rely more on greenhouse gazes-emitting electricity generation (e.g.: US - so including NY mentioned in TFA) : pollution of electric car is still a little bit cleaner than pollution of ICE cars.
- Only in countries with the most catastrophic power generation (e.g.: China, India, Australia, some african countries) are electric-car and ICE cars equivalently polluting.
1) Most power is still generated by burning fossil fuels. You're just moving your tailpipe elsewhere.
While at the same time, there are detailed analysis (google is your friend) concluding that:
- Even in countries that mostly produce electricity by burning fossils - like TFA's US - electric cars still have a better carbon foot print simply because this "moved away giant tail pipe" is much more efficient that the "original small one" attached to an ICE. In other words: yup, us power plant also produce carbon. BUT us power plants produce less CO2 per km than would an ICE, simply because the are optimised mostly for their efficiency, whereas the ICE is also optimised for size and for peak acceleration.
- In country that don't burn fossils that much for electricity (e.g.: lots of central or northern european countries), carbon foot prints are even better than the above.
- Very few countries like China, India, Australia and a few african countries (so nothing to do with NY) are so much reliant on dirty electricity that, electric cars and ICE don't differ that much in their emissions.
2) Cars don't contribute anywhere near as much to greenhouse gasses as we are led to believe. Cow farts are actually the #1 source. Buying basic goods made abroad also contributes significantly to the problem.
Yes the industry (including agricultural industy) also produce greenhouse gasses, and in bigger quantities than cars. That doesn't mean that you shouldn't use electric cars, that only means that you *should also* try to recude the industry's greenhouse gaz emission. (Which is beyond the point of TFA. Also, I don't know in the US, but several countries in Europe are also working toward lowering industry/agriculture emitted greenhouse gazes. e.g.: encouraging local grown food. So it's not as if electric cars were at the detriment of fight emissions in the industry)
3) The manufacturing of batteries in electric and electric hybrids is an incredibly dirty process. 4) The batteries in these cars aren't recyclable at all and will have degraded significantly after about 5 years of use. ICE cars don't need a new fuel tank every 5 years.
Yeah because ICE cars grow on tree. Organic trees.
Not.
Check detailed studies, most of the serious one also take the battery manufacture into account. So again, in anything but the few top most fossil heavy countries, the production of the battery is still offset by the reduced emissions while driving. That includes the US (even if it relies more on fossils than others and thus the advantage is less visible).
6) ICE cars can actually be carbon-negative. Boeing developed a workable method to grow ethanol in the world's deserts. Think about it: the Sahara Desert turned into a carbon sink. (Link: http://energypost.eu/exclusive...). With such a fuel, it would actually be better if everyone drove a gas guzzler until we meet an agreed upon level of carbon in the atmosphere.
Nice story, bro. But:
- It's still only one of those dozens "soon, new technology will make everything better..." over-hyped big media spin on some scientific advance. (I don't want to denigrate the scientific advance, but it takes a lots of steps between the lab and actual mass production). It's still only a story in some news paper. Whereas electric cars are currently available from several manufacturer. And low-emission electricity is also a reality in lots of countries.
Until Boeing starts producing and selling huge amounts of their bio-fuel, it makes sense to support electric cars.
- Carbon foot-print gets negative *only if* this production method consumes more carbon than the bio-fuel releases at the end. (e.g.: if only the fruit of a plant are used for fuel, and the rest of the plant stays keeping its carbon
On the other hand, google already managed something similar in the past : its page rank system.
Back when Google was simply a keyword search engine, it didn't simply return *all* webpages (that it knows off) where the query keywords appears. it did return *the top* webpages, using a whole ranking system to assess the quality of the page.
Whereas other more primitive search engines could be easily fooled by a link farming (e.g: forum and wiki spamming), it did require quite some art to manage a google bomb successfully, with the developers at google constantly refining their algorithm (pages with the same keywords showing up won't be given the same importance depending on their rank and/or the rank of pages leading to them and/or quality of the links).
The same here : instead of feeding the "featured snippets" AI with whatever the google crawler find, the snippets AI will eventually need to have the concept of "confidence level" associated with the information. Being able to react differently when the AI parses information from a reputable source (a peer-reviewed scientific article) (and/or even being able to autonomously process retraction of such source) and when the AI parse information on some "dubious" site (some extremist's site with an agenda).
That won't save google from a well though-out and coordinated google-bomb (snippet-bomb ?), but will at least avoid the AI blindly believing every bullshit it reads on-line. (On the other hand, given the gullibility of the average people, the current snippet AI isn't reacting much differently that this weird uncle that is always blabering about some conspiracy theory he read on-line somewhere. Successfully "passing turing test through stupidity" ?)
I take it you aren't old enough to remember Lotus 123?
I'm young enough to *have been conceived* more or less at that time. (And anyway, my parent were living on the other side of the Iron curtain back then).
OTOH when Lotus 123 hit? ZOMFG suddenly EVERYBODY had to have Lotus 123, from the smallest business to the biggest corp you weren't shit if you weren't using 123.
Though I couldn't be able to remember a period I didn't live through, I do have some general culture. Yup, I know that Lotus was one of the first massively successful spreadsheet software. It owes it to a few key thing, among the fact that everybody can see use of making some calculation (be it for business use as you mention, or any other use. Lab technicians love to have their protocols in the form of spread-sheet that can magically auto-adapt given a few key parameters), and it was relatively simple to use (as opposed to write your own data manipulation script, e.g. using the in-ROM embed BASIC) so that even non technical people could use it.
And it was very well marketed.
God that makes me feel ancient, but the reason your AI wouldn't do it is because it will always be a niche application as there is a very limited subset of people that are gonna actually WANT an AI on their PC, most will think its creepy or weird and say "do not want".
First, I'm not saying that AI as it is deployed NOW is a killer app. I'm only saying that is has a potential to soon evolve into a must have technology because it helps automate tons of tasks. (I used the word "candidate").
It's a bit like for some time Apps were just considered a "fad". But now some of them are slowly becoming "must haves", are driving hardware sales (you need a good enough phone to be able to use said app. Your old EPOC-powered Ericsson or Palm OS-powered PDA won't cut it, etc.) And a platform that isn't part of any large app ecosystem is probably going to day. (case in point : Microsoft's Windows for Phone only runs microsoft apps and never managed to break into the current Google/Apple duopoly) (counter example: Jolla's Sailfish OS. It's a faily small effort, barely known outside geeks. Beside being a fairly standard GNU/Linux distro for smartphones, also features various solutions to run Android apps. Its users aren't left aside. As such it has still managed quite a few success on the levels of manufacturer (Intex - the indian electronic devices manufacturer, not the US swiming pool manufacturer) or even states (Russia and China have shown interests into the platform). That's quite an achievement for such a small team).
My opinion is that even if currently it's only tech demo that creep the shit out of end users (Do you want me to automatically tag all your friends in all your facebook photos ? Do you want me to pre-write draft of your most likely answer on Google Allo ?) the technology has over-all good potential to be useful for end users (e.g.: better voice recognition and thus better dictation than the previous statistical methods used up until now).
The best part is that these are much more autonomous learning technology, relying less on smart algorithms to process the data, but more on the neural net simply "learning the data" and how to make sens out of it (Like a very young children that mostly "learns" the world around simply by experiencing it). Just like VisiCalc/Lotus 123/etc. made data processing much more accessible to people without programming skills, maybe AI could make bigdata analysis much more accessible to people who aren't necessarily good at developing subtle data processing algorithms and/or complex statistical methods.
Depending on how the technology is used maybe within a decade AI *will* be a must have that you can't do without. And if that becomes the case, you'll eventually need good hardware to be able to run the deep neural nets locally.
But indeed, nobody want the current AI that automatically recognise all your friends puking on your facebook photos.
The problem is, only Google gets to decide what goes in there, unless you go to the extensive effort of forking it and maintaining your own separate lineage of android-like OS. That might be uncomfortably too much control to one single US company.
On the other hand, Jolla's Sailfish OS is build on the "mer" core, a descendant of the Meego / Maemo that the same engineer were designing back when they were still at Nokia : it's a fairly standard GNU/Linux platform, with much more diverse contributors.
Is stuff like that still possible on Sailfish . . . ?
Basically: yes .
Jolla's Sailfish OS is based on a GNU/Linux base called "mer", a direct descendent of the Meego / Maemo that was featured on Nokia Nxxx devices. Actually written by the same exact people. (When Nokia left their R&D department, the engineer started Jolla and kept working on the same).
You can indeed run most of regular Linux software on it. There's even an "openbuild system" server that can produce appropriate RPMs for you (no need to locally compile them with the SDK) and a 3rd party repos for the RPMs. (openrepos, just like to good old Maemo days).
The only key difference is that they have switch their interface from GTK to Qt-Quick / QML. And though it's not released under a Free/libre license yet, you can still see the source and hack it due to the textual representation of QML.
- Skype has also a web interface. You can either run into your firefox browser (thus no local daemons). Or use Eion Robb's excellent "skypeweb" plugin for Pidgin / libPurple. (And last time I've read about it, this official new Linux client is basically the web app in its own self-contained app)
- OTR is a message encryption that works over any protocol. It's completely agnostic to the messaging system underneath. If you want to avoid your privacy exploded, you could try to use it (but if both of your use non standard client able to run OTR atop of it, you might better switch to another network).
Well, this one is not only "64-bit" only, it is specifically amd64 only. No arm64, no siree. And I'm typing these words on an armhf laptop.
Remember that Skype *also has* a web client. (And the latest Skype for Linux was basically that skype web-app warped together in a single-package together with google blink and a few key plugins).
Note: For Audio/Video calls, the web relies on ORTP, Microsoft's own Not-Invented-Here incompatible copy of WebRTC.
On the positive side, the JSON/XML protocol use by the Skype Web client has been successfully implemented as a Pidgin pluging by Eion Robb ("skypeweb"). Currently chat works (including group charts). Last time I've check, the developer had also being to look for supporting audio/video calls through the same interface as the skype web app.
Now if some new "killer app" like the next Lotus 123 comes up? This may change but so far I've seen nothing on the horizon that would fill that slot.
A possible candidate for such a "killer app" for more powerful CPU and GPU would be AI.
There's a crazy amount of interesting stuff that can be done by training deep neural nets to do your work. Some of this interesting stuff is more geared toward R&D that would run the workload on a high performance cluster anyway. But some of this has very practical application for everyday life (think RNN-LSTM in automatic natural speech transcription, think all the assistant such as Siri, Cortana, Alexa, OkGoogle, Soundify, etc.)
The problem, is that these neural nets have huge processing requirement to train, and even to use require some processing power. So currently, most of these AI assistants run remotely on the cloud. But by having more computing power locally could make such assistant more useable offline.
Why is no manufacturer capable or willing to release a smart phone with 100% free software from the start?
There *ARE* a few example of nearly 100% software smartphone.
Random examples: - OpenMoko's FreeRunner. (about a decade ago) Basically a "do the fuck you want to do with it" platform, which was designed from the ground up to run free-software. At least within what was possible back then. They manage quite well on some fronts (the GPS and the baseband are basically glorified external modems talking over a serial line with the main computers), and butchered a bit other elements (screen and GPU aren't quite correctly matched - the GPU isn't able to do accelerated 3D OpenGL on a buffer the size the screen has at native resolution) (and there was quite a scandal around some firmware that got stolen and reverse engineered breaching some NDA in the process) But at the end, it was a device were you could run lots of different things.
- Fairphone 2. The second device designed by Fairphone is on purpose designed to give possibilities to the end-users to replace the OS. AOSP is on purpose supported on the device. (I don't remember if you can order a phone with AOSP installed out of the box. But AOSP is among the officially supported OSes, and is installable as easily as possible for end-users) The device has an unlocked boot-loader and is designed to help other efforts to port OSes (there's a very functional port of Sailfish done by the community)
There is effort underway to make upstream kernels run on the hardware (currently the device is limited to an older 3.10 kernel due to hardware). Hope is to manage what RaspPi has (was initially locked-version kernel due to drivers, but latest kernels finally support all of the hardware)
If they really have to keep the baseband firmware proprietary, isolate it with an IOMMU. What are they afraid of?
There are a few stopper that prevent open-source software:
1. - Effort / End-user: the total number of geek that actively want an opensource OS on their phone is only a tiny fraction of the market of a smartphone. It's not worth actively spending resource to make sure that users can install opensource software on it (e.g.: OpenMoko was a small scale operation by geeks, for geeks). But, simply by selecting components that are at least supported up to some point might make the work easier. (That how Fairphone 2 is doing it now, that's how openmoko was built back then).
Though on the other hand, there might be a bigger number of end-users who would have a more general interest of not having their privacy completely raped by the crapware preinstalled on the phone - but they don't all understand that having a full FLOSS OS supported would help. (But: Fairphone pushed for AOSP exactly for this reasons : so privacy conscious users can have at least a possibility to have an opensource OS that can be reviewed by others. Turing phone is specially targeted at privacy conscious / security minded users. It comes with Sailfish OS pre-installed. Though it's not fully free/libre it's pretty close to it : the core is Mer (the descendant of Meamo/Meego : a pretty normal GNU/Linux base), and the interface is QML based (so although the license isn't currently free/libre, the source is directly accessible for patching being non-obscured/non-minified text. Jolla is promising to eventually completely opensource the whole Sailfish)
2. - Baseband As you mentioned, for licensing reason, the baseband firmware can't be opensourced. Only someone holding a cellular-band license can modify a software that control radio emitted on these frequencies. Now this comes with an even bigger problem : the baseband modem itself. For cost and space saving reasons, on most recent device the modem isn't a separate chip (that used to be the case back in the openmoko era, and back when TI's OMAPs where the most popular smartphone CPUs - the modem was logically a comp
There you'll see a *lot* of Linux workstations, and quite a few linux laptops (with the remaining being mostly Mac OS X laptops - also a variant of Unix, but at least the OS-and-hardware integration is done by one single company, removing a few headache of "how do I install Linux on this weird hardware ?")
Most popular distros here around seems to be : - Redhat / Fedora (because "enterprise") mostly on the workstations. - Ubuntu (because ultra-popular) mostly on the laptops - Mint for the "I don't want to be an Ubuntu sheep" crowd, specially those who vocally criticize Ubuntu. - Suse (for the people who actually want to be a real alternative to the Redhat and Ubuntu users, for the people who appreciate Suse stellar hardware support, for the people who appreciate Tumbleweed being a serious rolling distro made accessible for the non hard-core geek). - Gentoo (yup ! real people using gentoo in a production environment) for the hardcore geeks.
Slashdot Mirror
Ok.. It depends.... Password rules, like anything, when used within reason CAN increase security.
There has been some research which arrive at the conclusion that yes, indeed, password rules are actually bullshit for security.
As mentioned in the summary, enforcing password rules will actually block provably safe passwords :
- a base32 encoded 128bit pure random number. It's mathematically provable to be secure (if done by a cryptography-grade true random number generated, it's a 2^128 security, which is pretty good enough). But it's a 25 character long string of alaphanumeric. So it's not mixed case, and doesn't contain punctuation so it will be rejected by most stupid rules (also some rules have size specified as a range [9 to 16 characters], not a minimum [more than 8]. This will also reject a 25-long password).
As shown in presentations at numerous presentation in conferences such as CCC :
- even a complex rule set (Mixed case, must contain numbers and punctiation, at least 9 characters long) will usually give results such as "Denver17!"
Which are a lot less secure because they follow a general pattern (The first letter is the single capitalized, number come at the end, punctuation is the last and 9 out 10 times it's a '!' ). Most of these "rule abiding password" follow one of very few such patterns, and patterns are alarmingly easy to crack.
As such, no matter what, rules are a bad idea.
On the other hand, password managers with a generation function (like the above 128-bits equivalent password) are definitely a good idea.
WebEx is a horrible pile of java that isn't very portable (relies on binary JNI libraries for audio/video) and thus doen't work reliably on Linux 64 bits.
Adobe Connect is an aweful Flash monster (requieres freshplayer plugin layer under Linux) that crashes at the slightest disturbance (whenever the network devices or the mic sources change).
Meanwhile, all the various "meet"-type services (hangout meet, jitsi meet, etc.) rely on pretty standard HTML5/WebRTC. It works on anything under the sun that is compliant.
(Skype is in a middle ground. The older app might get problems when trying to run. And web Skype relies on ORTC, microsoft's own nih reinterpretation of the standards. So Microsoft Edge-only for now, but efforts are underway to bring it to linux).
It's also a good way to draw attention to yourself... very few people actually use it, so a search of just them is much more practical.
Depends on the region. Europe in general is much more privacy conscious. Specially some central / northern parts (E.g.: Germany, Switzerland).
Here you find people who use Tor regularily, just for the sake of keeping their life private.
You can thank these kind of people to help fill tor with background noise traffic.
Their are the one you make your reading of the anarchist cookbook less obvious
And who make "rounding up all tor users" completely impractical.
Planes become gliders when the engines fail.
(I might not be a pilote but I have a vague notion of the physics involved)
Just like helicopters become autogyros when their engine fails.
(also: as long as the engine didn't take the commercial airplane's wing with it when exploding)
(also: and some fighter jets have enough body lift [and avionics to compensate] to afford losing a wing)
I wasn't commenting on the merits of whole plane chute, just mentioning that these are being developed.
(But currently, the only reports I've came across are from the manufacturer themselves. So of course these are going to be heavily biased toward "Our products are incredibly useful and save tons of lifes !")
authorized USB sticks
Pay attention that the current default behavious of VBox scripts might open a different kind of vulnerability :
USB-pass-though requires that the VBox process has access to the raw USB device.
This is done by the script "/usr/lib/virtualbox/VBoxCreateUSBNode.sh"
it creates the appropriate entries in "/dev/vboxusb/"
granting them full group access for "vboxusers"
Currently this script is called by default by "/etc/udev/rules.d/90-vbox-usb.rules" for any plugged-in device.
That means the raw USB device of *any* USB gizmo is available for *ANY* process that runs with VBox's group credential.
This opens quite a big hole.
(Virtual Box it self then use a white list so *NOT ALL* device will be available to the Windows 10 VM, only those that you grant access to.
But it means potentially any USB device could be hacked by any process running with "vboxusers" group privileges).
A better way is to comment out the insersion rules, and only create the devices for device that you want to be visible to virtual box. :
Example of a configuration file that only grants access to a few of my devices that rely on windows-only software for firmware updates
# TomTom Live
ATTRS{idVendor}=="1390",ATTRS{idProduct}=="5454", ACTION=="add", RUN+="/usr/lib/udev/VBoxCreateUSBNode.sh $major $minor $attr{bDeviceClass} vboxusers"
# iRobot Scooba 450
ATTRS{idVendor}=="27a6",ATTRS{idProduct}=="0001", ACTION=="add", RUN+="/usr/lib/udev/VBoxCreateUSBNode.sh $major $minor $attr{bDeviceClass} vboxusers"
# UE MEGABOOM
ATTRS{idVendor}=="046d",ATTRS{idProduct}=="0a53", ACTION=="add", RUN+="/usr/lib/udev/VBoxCreateUSBNode.sh $major $minor $attr{bDeviceClass} vboxusers"
#SUBSYSTEM=="usb_device", ACTION=="add", RUN+="/usr/lib/virtualbox/VBoxCreateUSBNode.sh $major $minor $attr{bDeviceClass}"
#SUBSYSTEM=="usb", ACTION=="add", ENV{DEVTYPE}=="usb_device", RUN+="/usr/lib/virtualbox/VBoxCreateUSBNode.sh $major $minor $attr{bDeviceClass}"
SUBSYSTEM=="usb_device", ACTION=="remove", RUN+="/usr/lib/virtualbox/VBoxCreateUSBNode.sh --remove $major $minor"
SUBSYSTEM=="usb", ACTION=="remove", ENV{DEVTYPE}=="usb_device", RUN+="/usr/lib/virtualbox/VBoxCreateUSBNode.sh --remove $major $minor"
As suggested by other discussion threads here around :
You can also achieve the same virtually :
"virtually move" the image to an area without any signal.
I.e.: .iso files you mount) and shared folder (VirtualBox sharing doesn't go through network, so it's not opening windows 10 to remote access, at least not without a collaborating host OS).
Windows 10 goes into a VirtualBox VM.
VM has no network.
VM has only CD-ROM (so can read from
You can pass the files and necessary application through shared folders and .ISO respectively.
I would suggest avoiding USB pass-through because it's complex to configure it in a secure way (defaults VBox scripts just make all device visible to any application running with VBox's goup credentials).
You achieve the same as moving a Windows 10 laptop to an area without any signal.
(i.e.: No network for Windows 10, no matter what)
But you still get to have an actual connection on your host OS (say a well secured and well kept Linux host).
Commercial jets aren't indeed doing ejection seats.
But some *do* consider adding chute that could try to e.g.: save the cabin in care of dramatic loss of engines etc.
ICE cars, and their engines, are made from recycled and recyclable materials. {...} And again, Lithium Ion batteries are nothing but toxic landfill when they've outlived their ability to hold a meaningful charge.
And again, over the whole lifetime of the vehicle from manufacturing, through using, to decommissioning, the numbers in current report tend to show that even when factoring everything in, electric cars are still a bit (US) or some (EU) better. (Unless your country is China. Then it's basically the same).
(Don't forget that over their lifetime, ICE will release a significant of poluant too. It's a slow trickle, compared to decommissioning a car battery. But it builds up to very large total quantity. You just spread it more geographically and across time. Which might make it even more difficult to process).
And again, cars don't really make up a huge part of greenhouse gases.
Sorry, no. *THEY DO*. (I'm nitpicking at your usage of the word "huge" vs some superlative like "top contributor").
At 26% (EPA numbers) they are not the *biggest emitter* (that crown goes to power generation according to the same source).
But they are still a significantly big chunk of it. Significant enough they deserve to be considered and also addressed.
Our time and effort is better spent elsewhere.
That would have been the case if cars only contributed to say 4% of emissions.
That's not the case.
Again, even if they are not the *top* emitter, it's still worth spending effort on cars along with all the other emitter that still emits significant chunks of the pollution, even if they are not number 1 position on the ordered list.
Oh, and never mind that on a road trip you're stuck at a charge point for 30 minutes in order to recharge your car enough to drive another 2 hours.
Yeah! Let's drive 8 hours straight without a single break in it! Yay for safety! We human absolutely have no single problem to keep concentrating after hours of driving in a row!
Yup, maybe you're highly organized, doing all your road-trips in teams having at least 3 drivers and driving in shifts between drivers seat, passenger seat and sleeping in the back, while an AI supplement the passenger in their role of helping the driver watching the road.
Or you're one of those guys pretending that he can drive 12hours straight like a machine, only stopping to pee. Simply because up to now they happen by luck not to have had an accident. Even if there are massive study with numbers showing that this is a huge risk, actually.
As battery swapping stations haven't caught yet up, your and your co-driving friends' specific use case isn't well served by vehicles that can only rack up to somewhere around 400-500kms between 30-45min breaks. (And please tell me you're actually of the "taking shifts" variety, not the "Hey, I've driven 12hours only stoping to pee, and I'm still alive, so nobody ever needs to rest !")
You fail to take into account all the other uses cases :
- People who actually take 30min break each 2hours because that's the current recommendation to avoid being to distracted/tired/etc.
- People who actually drive less than 300km per trip (in fact, most of the trips done with cars tend to be under 100km in Europe. Even the cheapest battery option at my local car-sharing covers it without problems).
etc.
Finally, the tech described in the Boeing link isn't a "nice story, bro." It's real and Boeing is actively ramping it up.
But still, there's no gaz station where I can tank my car.
So for now, electric car are still having a better impact on the environment.
(Even when you factor everyting including the battery in).
That might change one day. But we still aren't there yet.
and the production method described does exactly what you describe -- use more carbon than it releases.
Well, that's good.
Couldn't you make a phosphorescent front projection screen?
In theory: yes, specially since there exist phosphors that only react to some laser frequencies (so you could even do RGB). (See Thunderf00t's Youtube video playing with blue phorphor balls and blue lasers).
You could in theory spent some resource in researching phosphors that reacts more or less to a laser, like older CRT's phosphor did react to its electron gun.
In practice: it's much cheaper to try to pump-up the frame rate and refresh the vector figure at a rate where the human eye's persistence of vision doesn't notice the flicker (much). (At least that's what I've read people are doing with OpenLase)
(And that's how the laser are used in practice in night clubs : drawing the vectors at very high refresh rate).
So now Dark Web means SMTP and FTP? Web? Web. *facepalm*
More seriously : Nope, not at all.
Dark Web has nothing to do with *what* you're talking to (usually websites, so usually HTTP/HTTPS),
but *how* you talk to (instead of opening a simple TCP connection between your laptop and a webserver, you send your request using some form of indirect complex hard-to-track routing. e.g.: TOR (The Onion Router)).
On the tor network, there are such things as tor servers which are accessed using a special address.
e.g.: Duck Duck Go has one.
From the outside, it doesn't look like your laptop is connecting to DDG's server.
It only looks like a bunch of tor nodes passing packets around.
At least in the case of Duck Duck Go, they also have an official DNS name that is registered to them, an IP address that can be backtraced to the server on the rack in the data-center where they are renting (or to the CDN that can then in turn further find them).
But some server only exist as obscure Tor keys. It's very hard to trace them in the real world.
Hence the name "Dark Web". The web of server that is hard to trace cause they don't work over plain vanilla TCP/IP.
It doesn't make sense that 85% of them suddenly stopped being greedy and slimey all of a sudden.
Every criminal thinks they will never get caught, or has already resigned themselves to the idea that it is part of the game. Some might have closed up shop, but 85%? I highly doubt it.
Not 100% of what goes on the dark web is criminal.
Take tor as an example:
Yes, some of them use .onion tor web services for the purpose of hiding ethically-dubious criminal activities.
(silk road used to be an example back then).
BUT, some use them for very practical reason like evading censorship (though it is *still* considered illegal in some specific jurisdiction, globally it's not and it's hardly unethical). .onion tor server.
The popular Duck Duck Go search engine also has a
If you're in China, and want to get informations about Tibet or the Tiananmen Massacre or simply about Taiwan, you can't through a normal search engine. They are either blocked in China or collaborate with the government and only report government-approved informations.
On the other hand, if you have access to the tor network (along with VPNs that's one of the few popular solutions to get around China's Great Firewall) you can search the web using Duck Duck Go's onion tor server. (And then read the source using your tor socks proxy).
As needs come and go, or because ressources are limited, some of these non-criminal dark web site are going to disapear.
I wonder if you could just bounce a laser (or simply a highly focused light) off a MEMS mirror or something. Or maybe you'd use multiples?
Such lasers are popular in nightclubs.
There are a few hacker who have adapted them to mame (google "openlase")
The main problem is that CRT Vector display have some persistance (the phosphores don't imediately turn black) and the image looks stable, whereas the laser make the image "blink" (unless you find a way to bump up the refresh rate)
My understand is that the NES system was based on timing of scan lines on the CRT
That's one of the two method that it can use.
Because the CPU of the NES is very slow, it's reported to be rather shitty on the horizontal axis, it's mostly good only for the vertical axis.
(And doesn't work at all with an LCD or a 100Hz CRT. It could at least work for the vertical axis on some laser scanning projectors, but forget about the horizontal axis, all these projectors ping-pong left and right)
Operation Wolf is reported by other to rely on this technique.
Most later consoles (basically anything until the Wii) counted exclusively on this method.
LCD-/100Hz- workaround used infrared light-sources and image processing (similar to how wii-mote works) and generate appropriate signal to simulate the position. (G1 Light Gun was produced for playstation this way).
The other method consist of quickly blink the screen and the various target black or white.
The game knows which target the gun is pointed at by knowing which targets were blinked white at the moment the gun detected a signal.
(e.g.: duck hunt turns first the screen black, then each duck is flashed white in turn. When the gun gets a signal it means it's pointing to the currently flashed duck)
in theory it should be able to somewhat work with a LCD, provided that all image post-processing is disabled and thus the LCD displays pixels as it is receiving them, so that the gun see exactly what the NES is trying to display.
(e.g.: when duck hunt first turn the screen black, the gun must imediately see black - so it know the light source it's pointing to is the screen. If there's a delay, the gun might still see light from the previous frame and decide that the player is trying to cheat by pointing the gun at a lamp)
With current processes for extracting Lithium from the ground resulting in 0.02% lithium and 99.8% dirt that is now contaminated by the toxic chemicals used to extract the lithium and the resource depletion on local water sources as water is shipped to lithium mines in salt flats, plus the fact that electricity to charge the cars likely comes from burning fossil fuels anyway, I'm going to guess the net environmental impact of an electric vehicle over the course of its life is only nominally better than a combustion engine.
According to studies (google is your friend) it all depends on the country.
- In countries that rely very little on greenhouse gazes-emitting power production (e.g.: lots of central and northern european countries) : pollution generated by the manufacture of battery and production of electicity is still a lot smaller than the pollution generated by the production of fuel and burning it in ICE.
- In countries that rely more on greenhouse gazes-emitting electricity generation (e.g.: US - so including NY mentioned in TFA) : pollution of electric car is still a little bit cleaner than pollution of ICE cars.
- Only in countries with the most catastrophic power generation (e.g.: China, India, Australia, some african countries) are electric-car and ICE cars equivalently polluting.
1) Most power is still generated by burning fossil fuels. You're just moving your tailpipe elsewhere.
While at the same time, there are detailed analysis (google is your friend) concluding that :
- Even in countries that mostly produce electricity by burning fossils - like TFA's US - electric cars still have a better carbon foot print simply because this "moved away giant tail pipe" is much more efficient that the "original small one" attached to an ICE.
In other words: yup, us power plant also produce carbon. BUT us power plants produce less CO2 per km than would an ICE, simply because the are optimised mostly for their efficiency, whereas the ICE is also optimised for size and for peak acceleration.
- In country that don't burn fossils that much for electricity (e.g.: lots of central or northern european countries), carbon foot prints are even better than the above.
- Very few countries like China, India, Australia and a few african countries (so nothing to do with NY) are so much reliant on dirty electricity that, electric cars and ICE don't differ that much in their emissions.
2) Cars don't contribute anywhere near as much to greenhouse gasses as we are led to believe. Cow farts are actually the #1 source. Buying basic goods made abroad also contributes significantly to the problem.
Yes the industry (including agricultural industy) also produce greenhouse gasses, and in bigger quantities than cars.
That doesn't mean that you shouldn't use electric cars, that only means that you *should also* try to recude the industry's greenhouse gaz emission.
(Which is beyond the point of TFA. Also, I don't know in the US, but several countries in Europe are also working toward lowering industry/agriculture emitted greenhouse gazes. e.g.: encouraging local grown food. So it's not as if electric cars were at the detriment of fight emissions in the industry)
3) The manufacturing of batteries in electric and electric hybrids is an incredibly dirty process.
4) The batteries in these cars aren't recyclable at all and will have degraded significantly after about 5 years of use. ICE cars don't need a new fuel tank every 5 years.
Yeah because ICE cars grow on tree. Organic trees.
Not.
Check detailed studies, most of the serious one also take the battery manufacture into account.
So again, in anything but the few top most fossil heavy countries, the production of the battery is still offset by the reduced emissions while driving.
That includes the US (even if it relies more on fossils than others and thus the advantage is less visible).
6) ICE cars can actually be carbon-negative. Boeing developed a workable method to grow ethanol in the world's deserts. Think about it: the Sahara Desert turned into a carbon sink. (Link: http://energypost.eu/exclusive...). With such a fuel, it would actually be better if everyone drove a gas guzzler until we meet an agreed upon level of carbon in the atmosphere.
Nice story, bro. But :
- It's still only one of those dozens "soon, new technology will make everything better..." over-hyped big media spin on some scientific advance. (I don't want to denigrate the scientific advance, but it takes a lots of steps between the lab and actual mass production). It's still only a story in some news paper.
Whereas electric cars are currently available from several manufacturer. And low-emission electricity is also a reality in lots of countries.
Until Boeing starts producing and selling huge amounts of their bio-fuel, it makes sense to support electric cars.
- Carbon foot-print gets negative *only if* this production method consumes more carbon than the bio-fuel releases at the end.
(e.g.: if only the fruit of a plant are used for fuel, and the rest of the plant stays keeping its carbon
On the other hand, google already managed something similar in the past :
its page rank system.
Back when Google was simply a keyword search engine,
it didn't simply return *all* webpages (that it knows off) where the query keywords appears.
it did return *the top* webpages, using a whole ranking system to assess the quality of the page.
Whereas other more primitive search engines could be easily fooled by a link farming (e.g: forum and wiki spamming),
it did require quite some art to manage a google bomb successfully, with the developers at google constantly refining their algorithm
(pages with the same keywords showing up won't be given the same importance depending on their rank and/or the rank of pages leading to them and/or quality of the links).
The same here : instead of feeding the "featured snippets" AI with whatever the google crawler find, the snippets AI will eventually need to have the concept of "confidence level" associated with the information.
Being able to react differently when the AI parses information from a reputable source (a peer-reviewed scientific article) (and/or even being able to autonomously process retraction of such source) and when the AI parse information on some "dubious" site (some extremist's site with an agenda).
That won't save google from a well though-out and coordinated google-bomb (snippet-bomb ?), but will at least avoid the AI blindly believing every bullshit it reads on-line.
(On the other hand, given the gullibility of the average people, the current snippet AI isn't reacting much differently that this weird uncle that is always blabering about some conspiracy theory he read on-line somewhere. Successfully "passing turing test through stupidity" ?)
I take it you aren't old enough to remember Lotus 123?
I'm young enough to *have been conceived* more or less at that time. (And anyway, my parent were living on the other side of the Iron curtain back then).
OTOH when Lotus 123 hit? ZOMFG suddenly EVERYBODY had to have Lotus 123, from the smallest business to the biggest corp you weren't shit if you weren't using 123.
Though I couldn't be able to remember a period I didn't live through, I do have some general culture.
Yup, I know that Lotus was one of the first massively successful spreadsheet software.
It owes it to a few key thing, among the fact that everybody can see use of making some calculation (be it for business use as you mention, or any other use. Lab technicians love to have their protocols in the form of spread-sheet that can magically auto-adapt given a few key parameters), and it was relatively simple to use (as opposed to write your own data manipulation script, e.g. using the in-ROM embed BASIC) so that even non technical people could use it.
And it was very well marketed.
God that makes me feel ancient, but the reason your AI wouldn't do it is because it will always be a niche application as there is a very limited subset of people that are gonna actually WANT an AI on their PC, most will think its creepy or weird and say "do not want".
First, I'm not saying that AI as it is deployed NOW is a killer app.
I'm only saying that is has a potential to soon evolve into a must have technology because it helps automate tons of tasks. (I used the word "candidate").
It's a bit like for some time Apps were just considered a "fad". But now some of them are slowly becoming "must haves", are driving hardware sales (you need a good enough phone to be able to use said app. Your old EPOC-powered Ericsson or Palm OS-powered PDA won't cut it, etc.) And a platform that isn't part of any large app ecosystem is probably going to day.
(case in point : Microsoft's Windows for Phone only runs microsoft apps and never managed to break into the current Google/Apple duopoly)
(counter example: Jolla's Sailfish OS. It's a faily small effort, barely known outside geeks. Beside being a fairly standard GNU/Linux distro for smartphones, also features various solutions to run Android apps. Its users aren't left aside. As such it has still managed quite a few success on the levels of manufacturer (Intex - the indian electronic devices manufacturer, not the US swiming pool manufacturer) or even states (Russia and China have shown interests into the platform). That's quite an achievement for such a small team).
My opinion is that even if currently it's only tech demo that creep the shit out of end users (Do you want me to automatically tag all your friends in all your facebook photos ? Do you want me to pre-write draft of your most likely answer on Google Allo ?) the technology has over-all good potential to be useful for end users (e.g.: better voice recognition and thus better dictation than the previous statistical methods used up until now).
The best part is that these are much more autonomous learning technology, relying less on smart algorithms to process the data, but more on the neural net simply "learning the data" and how to make sens out of it (Like a very young children that mostly "learns" the world around simply by experiencing it).
Just like VisiCalc/Lotus 123/etc. made data processing much more accessible to people without programming skills, maybe AI could make bigdata analysis much more accessible to people who aren't necessarily good at developing subtle data processing algorithms and/or complex statistical methods.
Depending on how the technology is used maybe within a decade AI *will* be a must have that you can't do without.
And if that becomes the case, you'll eventually need good hardware to be able to run the deep neural nets locally.
But indeed, nobody want the current AI that automatically recognise all your friends puking on your facebook photos.
Of course it's technically feasible. It's called universally agreed upon standards and conventions.
And webapps are slowly exactly becoming this.
Android is open source under the Apache 2.0 license
The problem is, only Google gets to decide what goes in there, unless you go to the extensive effort of forking it and maintaining your own separate lineage of android-like OS.
That might be uncomfortably too much control to one single US company.
On the other hand, Jolla's Sailfish OS is build on the "mer" core, a descendant of the Meego / Maemo that the same engineer were designing back when they were still at Nokia :
it's a fairly standard GNU/Linux platform, with much more diverse contributors.
Is stuff like that still possible on Sailfish . . . ?
Basically: yes .
Jolla's Sailfish OS is based on a GNU/Linux base called "mer", a direct descendent of the Meego / Maemo that was featured on Nokia Nxxx devices.
Actually written by the same exact people. (When Nokia left their R&D department, the engineer started Jolla and kept working on the same).
You can indeed run most of regular Linux software on it.
There's even an "openbuild system" server that can produce appropriate RPMs for you (no need to locally compile them with the SDK) and a 3rd party repos for the RPMs. (openrepos, just like to good old Maemo days).
The only key difference is that they have switch their interface from GTK to Qt-Quick / QML.
And though it's not released under a Free/libre license yet, you can still see the source and hack it due to the textual representation of QML.
2 key things :
- Skype has also a web interface. You can either run into your firefox browser (thus no local daemons). Or use Eion Robb's excellent "skypeweb" plugin for Pidgin / libPurple.
(And last time I've read about it, this official new Linux client is basically the web app in its own self-contained app)
- OTR is a message encryption that works over any protocol. It's completely agnostic to the messaging system underneath.
If you want to avoid your privacy exploded, you could try to use it (but if both of your use non standard client able to run OTR atop of it, you might better switch to another network).
Well, this one is not only "64-bit" only, it is specifically amd64 only. No arm64, no siree. And I'm typing these words on an armhf laptop.
Remember that Skype *also has* a web client.
(And the latest Skype for Linux was basically that skype web-app warped together in a single-package together with google blink and a few key plugins).
Note: For Audio/Video calls, the web relies on ORTP, Microsoft's own Not-Invented-Here incompatible copy of WebRTC.
On the positive side, the JSON/XML protocol use by the Skype Web client has been successfully implemented as a Pidgin pluging by Eion Robb ("skypeweb").
Currently chat works (including group charts).
Last time I've check, the developer had also being to look for supporting audio/video calls through the same interface as the skype web app.
Now if some new "killer app" like the next Lotus 123 comes up? This may change but so far I've seen nothing on the horizon that would fill that slot.
A possible candidate for such a "killer app" for more powerful CPU and GPU would be AI.
There's a crazy amount of interesting stuff that can be done by training deep neural nets to do your work.
Some of this interesting stuff is more geared toward R&D that would run the workload on a high performance cluster anyway.
But some of this has very practical application for everyday life (think RNN-LSTM in automatic natural speech transcription, think all the assistant such as Siri, Cortana, Alexa, OkGoogle, Soundify, etc.)
The problem, is that these neural nets have huge processing requirement to train, and even to use require some processing power.
So currently, most of these AI assistants run remotely on the cloud.
But by having more computing power locally could make such assistant more useable offline.
Why is no manufacturer capable or willing to release a smart phone with 100% free software from the start?
There *ARE* a few example of nearly 100% software smartphone.
Random examples :
- OpenMoko's FreeRunner. (about a decade ago)
Basically a "do the fuck you want to do with it" platform, which was designed from the ground up to run free-software. At least within what was possible back then.
They manage quite well on some fronts (the GPS and the baseband are basically glorified external modems talking over a serial line with the main computers), and butchered a bit other elements (screen and GPU aren't quite correctly matched - the GPU isn't able to do accelerated 3D OpenGL on a buffer the size the screen has at native resolution) (and there was quite a scandal around some firmware that got stolen and reverse engineered breaching some NDA in the process)
But at the end, it was a device were you could run lots of different things.
- Fairphone 2.
The second device designed by Fairphone is on purpose designed to give possibilities to the end-users to replace the OS.
AOSP is on purpose supported on the device.
(I don't remember if you can order a phone with AOSP installed out of the box.
But AOSP is among the officially supported OSes, and is installable as easily as possible for end-users)
The device has an unlocked boot-loader and is designed to help other efforts to port OSes (there's a very functional port of Sailfish done by the community)
There is effort underway to make upstream kernels run on the hardware (currently the device is limited to an older 3.10 kernel due to hardware). Hope is to manage what RaspPi has (was initially locked-version kernel due to drivers, but latest kernels finally support all of the hardware)
If they really have to keep the baseband firmware proprietary, isolate it with an IOMMU.
What are they afraid of?
There are a few stopper that prevent open-source software :
1. - Effort / End-user :
the total number of geek that actively want an opensource OS on their phone is only a tiny fraction of the market of a smartphone. It's not worth actively spending resource to make sure that users can install opensource software on it (e.g.: OpenMoko was a small scale operation by geeks, for geeks).
But, simply by selecting components that are at least supported up to some point might make the work easier. (That how Fairphone 2 is doing it now, that's how openmoko was built back then).
Though on the other hand, there might be a bigger number of end-users who would have a more general interest of not having their privacy completely raped by the crapware preinstalled on the phone - but they don't all understand that having a full FLOSS OS supported would help. :
(But
Fairphone pushed for AOSP exactly for this reasons : so privacy conscious users can have at least a possibility to have an opensource OS that can be reviewed by others.
Turing phone is specially targeted at privacy conscious / security minded users. It comes with Sailfish OS pre-installed. Though it's not fully free/libre it's pretty close to it : the core is Mer (the descendant of Meamo/Meego : a pretty normal GNU/Linux base), and the interface is QML based (so although the license isn't currently free/libre, the source is directly accessible for patching being non-obscured/non-minified text. Jolla is promising to eventually completely opensource the whole Sailfish)
2. - Baseband
As you mentioned, for licensing reason, the baseband firmware can't be opensourced. Only someone holding a cellular-band license can modify a software that control radio emitted on these frequencies.
Now this comes with an even bigger problem : the baseband modem itself.
For cost and space saving reasons, on most recent device the modem isn't a separate chip (that used to be the case back in the openmoko era, and back when TI's OMAPs where the most popular smartphone CPUs - the modem was logically a comp
Unix is king in some scientific domains.
There you'll see a *lot* of Linux workstations, and quite a few linux laptops
(with the remaining being mostly Mac OS X laptops - also a variant of Unix, but at least the OS-and-hardware integration is done by one single company, removing a few headache of "how do I install Linux on this weird hardware ?")
Most popular distros here around seems to be :
- Redhat / Fedora (because "enterprise") mostly on the workstations.
- Ubuntu (because ultra-popular) mostly on the laptops
- Mint for the "I don't want to be an Ubuntu sheep" crowd, specially those who vocally criticize Ubuntu.
- Suse (for the people who actually want to be a real alternative to the Redhat and Ubuntu users, for the people who appreciate Suse stellar hardware support, for the people who appreciate Tumbleweed being a serious rolling distro made accessible for the non hard-core geek).
- Gentoo (yup ! real people using gentoo in a production environment) for the hardcore geeks.