There's no pinout nor SPI difference between the 2. The only difference is a logical one.
SDHC cards come pre-formatted with FAT32. SDXC cards come pre-formatted with exFAT, and Microsoft has patented the shit out of it. So unless the company has paid money to Microsoft, they can't use exFAT and can only advertise "up to 32GB SDHC cards".
But nothing prevents you to buy a 128GB SDXC and : - either install a FUSE-exFAT driver on your OS if supported. - or reformat the card with something supported by the OS (depending on the OS: FAT32, F2FS, Ext3/4, BTRFS, UDF, etc.)
So 128/256/512/1024GB will work on most SDHC readers (i.e.: that support more than 4GB plain- SD), but the manufacturer can't advertise it because they lack the patents to the file format that is mandatory to advertise SDXC support.
Oh, you plug it in and it works, but you don't get all the fancy per-game remapping and 30s startup time and 1-2GB RAM consumption.
It works in my case.
The whole stack support them: - Linux' USB HID driver perfectly supports all the available axis and buttons my the MX Master. - The SDL library used as a middle ware support multiple extra mouse axis and buttons. - Quake3 supports binding any commands to any input. (You just get fancy name like "Right Click" on the first few buttons. Then it's more like "M_BTN7"...)
For stupid games that don't have a good re-mapper, there's always things like (rejoystick, joy2key, qjoypad, etc.)
As usual, the crap software that comes with your hardware is crap. (like the fancy stupid DVD player that came with your optical drive. Simply use VLC)
Most large companies have their very own copy of github in-house. Most commonly these days it's on AWS or some other cloud offering that the company controls for their IT Projects.
I can confirm.
Though we aren't *companies*, most of the universities and research institutes here around (Switzerland) have their own in-house git repository. Though in our case, a self-hosted copy of *Gitlab* is what is the most popular here around. And most of the time it's hosted on the universty's/research institute's own server because of complex IP/publishing/secrecy considerations
And if we do it, I can clearly imagin that huge corporation could be doing it too.
And for the record, Google has announced that they've shut down Google-Code for 3rd party project only. Android is *still* officially hosted on their servers.
Their flagship mouse MX Master has quite a few driver issues and also the DarkField sensor seems to have problems with some surfaces.
Drivers? For a *mouse* ? It's fucking USB HID device (or bluetooth wireless if your laptop has Bluetooth 4.0 LE and you go dongle-less). It's just a bunch of axis and buttons (though you have to admit that the MX Master mice tend to have a little bit more than the average of them). You just plug into anything with generic USB HID support and it should work out of the box. If it doesn't work, you shouldn't blame Logitech, you should blame the retarded operating system that fails at basic USB HID.
The only thing that would require extra software would be:
- Battery status in your status bar. Which is visible thanks to green LEDs on the mouse if self, so I doubt it's that much necessary to have the status-bar widget. (Though we're on/. Maybe you have a script that automatically sends an e-mail to your smartwatch to remind to plug the charger into the mouse when leaving the room)
- If you want to do some complex re-paring (e.g: pairing the mouse with a Logitech dongle that wasn't initially designed for it). (Again, this is/. and this might actually be the case. I would point you in the general direction of Solaar)
- Circumventing a piece of software that has asinine key/button binding. (e.g.: that has some stupid arbitrary hardcoded maximum of 3 mouse button, thus preventing you from binding the extra butons to additional functions). But then, any keyboard/mouse/joystick button remapping software would be plenty enough.
Logitech Quality has declined measurably recently, that's my experience anyway.
I have a strong impression that Logitech doesn't the same build quality in all markets. (I've noticed difference in products between US market, and Logitech's home Switzerland)
Their next iteration (Pebble 2 / Pebble Time 2) is *still* only bluetooth connected and with limited low-power processing.
But if you want to go smartphone-less, instead of trying to pack extra smartphone functionality into the watch itself (3G connectivity, Wifi support, whatever...) they also develop the Pebble Core a small screen-less android powered wearable computer.
Essentially, the screen-less iPhone of your joke, but for real, and powered by Android instead of iOS (and not called the "iPhone Shuffle").
Yes, we should eliminate all those things which are working great, so there are no ports, because... reasons
He just gave you a reason, smartass: so you can bring the phone underwater.
I have quite a few gadgets that can go underwater (not just splash resistant, but water resistant: full IPX7-certified immersion capable) that do feature ports.
Including the e-Book I gave present to my mother (Kobo H20) : that has ports - a microUSB charging/sync port, and a microSD card port. (As long as the rubber cover is over them, and isn't currently open to swap the microSD card, this thing can survive 1m under water for half an hour)
That also includes my speaker (UE Megaboom): it has a microUSB and a 3mm Audio stereo analog jack and a huge honking battery rated at a whooping 200Wh. Everything that Apple deems necessary to remove just to make their phone splash resistant. And this speaker is not only resist spalsh but can be forgotten for some time at the bottom of my above-ground pool. And will keep playing music (though not very audible due to difficulty of transferring sound air->water->air again).
I gave also a photo camera as a present to my girl (Olympus TG-4) this thing is rated at 15m depth (you could go diving with it). And it still has a USB connector, a SD card port, and Audio/Video connection (though the port is digital : it's a micro HDMI, not analog jack port) and fast and easy end-user replaceable battery. Though for the record it use plastic doors instead of rubber covers to protect the ports.
So again, can you explain me why Apple needs to remove functionality that is used everywhere (just go to any student party and watch how often people plug their smartphones with their music playlists into the analog jack of whatever speaker system is laying around*) just in the name of making the damn thing splash resistant ? Oh yeah, I know: their obsession of making the phone thin enough to slice cheese. A rubber port cover would stay in the way.
(It's the same company that obsesses with ridiculously small SIM card formats, even if the whole device isn't fitting directly in the ear - Lt.Uhura's style)
---
(*) Apple removing the audio jack from their hardware would be like removing the floppy drive from their "canddy" iMacs... before USB flash drive got available. Currently only some portable speaker system feature bluetooth. Even less are hassle free (NFC-based "touch to connect"). There are tons of old speaker systems in student dorms only accepting analog inputs. (Or even more ironically: accepting audio from the old (non-lightning) Apple connector that isn't produced anymore).
Luckily Bluetooth to 3mm jack / -to stereo RCA (and most ironic: to Apple 30-pin) audio dongle are cheap from China over ebay, so eventually this situation might get more jack-less friendly.
This is the sound piece in question >> It's called "plim plim" in Brazil.
And that sounds very close to one of the alternative chimes available on PalmOS devices by Palm Inc. (was it the "SciFi" chime ? I'm not sure anymore), available at least since the early 2000s (on my Palm IIIc back then).
Probably other PDAs have use a similar chime even earlier (could some e.g. Psion / EPOC user confirm ?)
Though trademark law doesn't have a concept of "prior art", it does have something similar to "defend or lose it". Currently this sound isn't distinctly associated with their branding (at least outisde of Brazil. Maybe there this "plim-plim" is immediately associated with the brand and is never heard anywhere else). It's not a jingle that is specific to grupo globo, it's something that has been heard around countless times. It's completely generic and the EU is completely right at laughing them out of the court.
I got 2 people in my group of friends (which isn't that big, those 2 are pretty much my closer circle of friends) who are anything but geeks (one a mechanic, one an artist) who BEG me to show them how to get away from Windows 'cause they don't want to deal with Win10 and all they do with their computers is pretty much playing games (and watching porn, of course). Now, doing the latter isn't the problem, but I can NOT at the current moment recommend to a non-technical person playing his games in Linux.
Current best bet are: - follow closely what Valve is doing (both in term of software: Steam, and SteamOS - and in term of hardware: Steam controller) - stick for now to a distro that is immensely popular: Ubuntu would be a somewhat good bet for now. - stick to hardware that is immensly popular, and check online forums, etc. - AMD since the Polaris generation is safe for out-of-box. - check games that use an engine that is well ported on Linux (something popular like Unity / Unreal / Cryengine / older idTech / etc.) rather than a game that use a homegrown engine that is only rushed to linux, or use actually a bad-quality wrapper. - Phoronix is a good place to get info about hardware, etc. their forums is also a possible place to ask around for help.
The only manufacturer still making hardware with 3D Audio is Creative and they do put some half baked broken OpenAL drivers that might work when it doesn't crash everything (thanks for the effort, Creative... next time just release the docs to dev and throw some salary) Nearly everything else is actually plain simple DAC hardware that is 100% handled by Asla and pulse (because there's no actual 3D hardware or anything. it just converts a 44100 or 48000 bitstream into analog signal). All the "3D Audio" is actually entirely done in software, often using an official service (I think DirectSound is the official one in Windows ?) and linux *does have* an OpenAL implementation, or much more likely entirely done in the game's engine.
Currently, linux dev have done all the possible. It's now up to: - Creative to fix their shit or release XFi docs and pay a kernel developer. - Game developer to try to leverage what is available (generally; either use OpenAL, or have the game engine do the processing and output to SDL which will handle the back-end interaction ALSA/Pulse)
and support for any kind of gaming hardware, from macro keyboard and multi-button mice to racing wheels to flightsticks to VR-hardware, is practically non-existing.
Razr is releasing docs (and I think giving donations to developers ?) so there's support underway, for all features including programmable coloured lights. (But this has started only recently, so I have no idea, how far is the out-of-box experience).
Logitech is sometimes donating hardware to developers and thus lots of their hardware is working, including advanced features like force-feedback. But it depends on models, so sadly you currently also need to track down which exact model before buying.
Most of the hardware is generally still working through generic HID drivers that will expose simply a collection of buttons/axis, so they are still somewhat usable even without the advanced features (force feed-back, special macros). But than the problem is again the game developers: they lack specially in the button assignment department, most games can't even assign buttons correctly and refuse to work with anything that isn't an Xbox controller.
And I tend to use a pretty simple setup when I play (regular mouse + keyboard, or regular joystick) so I don't pay attention to all the "macro" craze and thus can't give much more informations.
My general recommendation would be, in order: - use something actually developed with Linux in mind and with strong software support, e.g.: Steam controller - use simple vanilla controllers (keyboard + mouse, or joystick/wheels that work with basic HID). - use something with active development: Razr and check hardware support (it has only started recently) - use something with active development: Logitech and check hardware support (not everything is supported) - use something widely popular like an Xbox controller: Microsoft hasn't moved an inch, but the thing is so popular that there is support (just check).
Look at what walled-off, closed atrocity Apple has made out of Unix, and the people buy it. Why? 'cause it works.
Fixed that for you. Apple also runs an Unix, but it is based on BSD kernel (running atop of a Mach micro-kernel) instead of Linux. Very probably for licensing reasons (BSD vs, GPL).
Using CSS drivers on an equally proprietary piece of hardware is a VERY small price to pay to get more people to join the ranks of Linux users.
The problem is that you need *to be able* to run said CSS drivers. And that's where things get problematic. You see, Linux is a different beast than Windows where people still run the same Windows XP 15 years later.
Linux kernel get improved/fixed/modified/modernized. Some API/ABI change.
For opensource drivers, that are part of the kernel source tree, and that get dedicated developer to maintain (like those on AMD's payroll), that means that the driver will get also upgraded along and they will work with whatever latest vanilla kernel you get.
For closed source drivers, that means that they get regularly broken, and the hardware maker that wrote the driver need to get their shit done *ON TIME* so the closed source driver also works with Linux Kernel 4.8
And that's the problem here. Nvidia closed source work atop of a shim that they write themselves and that needs to be modified to work with every single last kernel version out there.
So currently your only solution is to run Ubuntu to which Nvidia happen to deliver preview version of drivers matching the kernel version of the distro.
For everything else, the situation sucks, and "nouveau" is the only remaining opensource driver that will be more or less updated and will at least load on any kernel version.
Contrast the situation with AMD: Since at least GCN 1.2 (previous generation of Radeon hardware - also doable with GCN 1.0 and 1.1, but currently not already mainstream), all drivers stack run atop the same kernal driver. This driver is opensource. This driver is in the main kernel tree and thus is available in any vanilla kernel. This drivers gets maintainers/developers some even paid by AMD themselves. Thus even on the latest kernel (4.8) you get a Radeon kernel module (amdgpu.ko) that actually works.
With the current generration of card (Polaris) on launch day, you could use it out of the box on your latest distro, it works. - you either use the stock opensource stack, that uses Mesa/Gallium3D to provide OpenGL 4.3 (and in the case of Polaris, that has performance close to the closed source offer) - or you swap the dynamic library providing openGL with AMDGPU-Pro, the official closed source openGL which runs atop *the same kernel driver, so also out-of-the box on the latest kernel version*, which will provide OpenGL 4.5 (and all the weirdness necessary for CAD software).
This is something you can buy and use *right now*.
Intel, although not playing in the same field, *also* provides launch day opensource support in upstream kernel.
Why couldn't Nvidia bother to move their ass and do the same? Oh way, that would require them to develop an actual linux driver, instead of simply recompiling their Windows code on linux and use a shim to pass along whatever is needed. (And thus regularly breaking/missing some corner case feature that aren't exactly handled the same way on linux... cough.. Optimus!...cough...)
Because more people using Linux means more reasons for hardware makers, DESKTOP HARDWARE makers, to put out sensible support for their hardware.
But Linux developers currently don't have the resources to do the hardware makers' job. If there's no launch day driver available, because closed-source code need to get upgraded and/or because there are no developer save a few volunteers, nouveau is all you could get.
- Either you put up with nouveau until Nvidia does their shit and then you do
Well, at least, some of the earliest casualties in the "Knowledge" part - those who worked on the Linux effort with Maemo/Meego - have since then founded Jolla and had some relative success in making a good full-blown GNU/Linux OS for phones with Sailfish OS.
(Happy user of a Jolla1 phone, looking forward for an official port to Fairphone2 including the Android Apps support bit).
The problem is that the kernel is always progressing.
So for the GeForce binary blob to work, you need to have an upgrade that works with the kernel version provided to you by your distro, which very often IS NOT available at the time of release.
Since Nvidia have discontinued their own 2D only "nv" driver, "nouveau" is the next best thing that is available in a modern kernel out-of-the box.
But that thing is developed by volunteers only, with nearly no help from Nvidia themselves.
So you're basically stuck to using only offical builds of Ubuntu, because at least Nvidia provide preview version of drivers in a timely manner.
Compare the situation with AMD, where Mesa/Gallium3D *AND* the closed source AMDGPU-Pro libraries both use the same opensource kernel module, that is upstreamed (its source is in the official kernel tree) and whose developpers include people on AMD's payroll: - currently, you know that modern cards that you buy now (e.g.: the current Polaris) work out of the box with any modern vanilla kernel. - drivers just boils down to swapping with dynamic library will provide OpenGL. Current Mesa 12.1 provides you with OpenGL 4.3 out of the box, or you could swap for the AMD's closed source AMDGPU-Pro library (which sits attop the SAME kernel driver which, as said, works out of the box) which will provide you the current latest OpenGL (4.5) and include all the weird quirks that you might need if you're working with CAD software. - forgot to mention: on official launch day, performance of both Mesa and AMDGPU-Pro libs were reported to be relatively close.
That's the kind of dedication I'm willing to pay for.
Polaris is out there now. Mesa 12.1 works with it right out of the box at OpenGL 4.3 level. So unless you're on some ancient distro (e.g.: you're locked on some LTS cycle for version-stability reasons) you can have it right now. (otherways you wait until *your distro upgrade*)
I'll just stay with what always just works for me under Linux: Nvidia.
You mean this piece of shit in my laptop which was a nightmare to upgrade with the official drivers (always needing to wait until Nvidia finally release an upgrade to the kernel driver to follow evolution of the upstream kernel) with very slow pace (my laptop isn't covered by the latest drivers anymore, I need to wait for the older legacy driver to get updated too) and that on occasion has massive graphical corruptions ? (And that's when the text console doesn't outright stop working ?)
And thus is completely unusable on any rolling-release distro (e.g.: Opensuse Tumbleweed ?)
Whose opensource drivers are developed by a bunch of volunteer to which Nvidia gives no documentation nor any other help except once every blue moon when they throw a bone with some Tegra-related docs that also happen to miraculously help the GeForce efforts too ? (And thus doesn't handle the blanking of my laptop ?)
Sorry, but no thanks.
Nvidia is unusable on anything that isn't : - a standard desktop with a generic GeForce card - or evetually a very recent laptop with a very classical build - running the latest official Ubuntu.
Meanwhile Radeon card gets docs and salaries by AMD. Latest Polaris is a smooth "out-of-the-box" experience. And in the past, the opensource drive has always been a smooth experience on the various build I've used it (though, I need to conceed, good OpenGL levels (e.g.: 4.x) have only been achieved in the recent iterations of Mesa).
I can easily imagine a lot of affected Note 7 owners -- approximately 2.5 million of them -- weren't exposed to the recall message for some reason.
We're speaking about smartphones.
- As per the various laws that were put into action in the name of "protection against terrorism" after 9-11, lots of countries (including the US) require tracking off all users of cell phone services (even pay as you go) to be properly registered. That means, at some point during purchase, this smartphone was associated with a SIM card that has an identity linked to it. It's not necessarily the SIM card that finally ended up in the phone (e.g.: the grand parents might have bought it, but kept their old phone with the SIM in it, and might have given the smartphone as a present to the 6yo kid), but easy to track down that the grandparents were present at the moment when the phone was bought. Through this way it is possible that the store chain that sold the phone (or the company that is keeping track of the cell phone customer for them) could track every last buyer of the phone and tell them. (They won't be able to tell the kid, but at least they can reach the grand parents who bought it).
If the way SIM-card holder registry is done in the US can't be used to track down users in case of emergency, then it's really a piece of shit that can't even stand up for the reason it was invoked (though I strongly suspect that this is the case. Just another privacy invading policy that can't be actually used against terrorism or for emergencies).
- Due to the way radio-wave licensing works, there are part of the phone radio circuits (parts of the firmware running in the modem, and the firmware running in the SIM card) that aren't update by Google / Samsung (they aren't part of the Android installation installed on the smartphone). They are not supposed to be end-users accessible. Only holder of license for the radio channel can access it. It's the carrier's job to send over-the-air updates to this part of the modem firmware, and to the SIM card. (When not straigh sending a new physical SIM card, because after 10 years the previous one is too old and useless to run anything compatible... been there...) (Also hope that your phone's SoC doesn't use it's modem chip *as a Northbridge/main controller* and the modem-firmware update causes freezes/crashes in the OS... cough... Qualcom!...cough...) Thus at any time every cell tower knows exactly which phone (and exactly which *hardware model* of phone) is connected to it (and might send such OTA updates). Cell tower can also send not only SMS messages, but direct service message (a full screen text/warning that bypasses the SMS system). Also keep in mind that, even without a SIM card, a smartphone *will* establish contact to a cell tower (at least to provide emergency calling services).
Thus it's trivially possible (with a bit of cooperation between carriers) that a message is delivered straigh to the screen of all concerned smartphones (i.e.: a specific IMEI range, or a certain firmware revision...) Even if the 6yo can't read the "WARNING: this phone is about to explode" message, he'll probably complain to the nearest adult that the pop-up is disturbing his session of Pokémon.
As a comparison, 1 year ago, here around a couple of store chains needed to recall quick-build above-ground inflatable spas/whirlpool due to potential electric shock hasard in the heating system. It's not even an IoT connected thing, there's no automatic way to remotely contact users / display warning on the console. It was entirely done through public communication / announcement. Still, the whole defective batch was successfully recalled without any causality. But it might be that germanic speaking countries might be a little bit better organised than the US.
Still, if they can manage to recall a dumb spa, whereas Samsung and the US cell carriers are unable to recall a connected device that can remotely display message and whose buyers are supposed to be easily traceable (for terrorism/emergency reasons), then there's a deep problem somewhere.
There's a lot less likeliness of Moonquales happening:
- The Moon, as mentioned by others is tidally locked with Earth. i.e.: It's always the same side facing Earth that means that Earth tidal stress is always "pulling" the exact same part. thus no change in this "pull" and therefore no directly-earth-caused tides on the moon.
- The Moon is a huge solid rock, it doesn't move much, whereas the Earth is pieces of solid crust all covering a molten mantle on which they more or less float (the "more of less" part being when they bump into each other, rub against each other, or one dives under another). Thus Earth has a tectonic seismology (the tectonic plate doing their above mentioned shit instead of peacefully floating next to each other), on which a whole bunch of factors (including in parts the tidal stress exerted by the Moon as maybe suggested by this study) can act as trigger to release to stored shear stress (causing Earthquakes). Whereas the Moon, by being a solid block of rocks can't do that.
- On the other hand, the Moon has no atmosphere : Any falling rock will smash straigh on the surface. Such meteorite contribute heavily on the moonquakes. The part of the moon that face the Sun will be backed in light, while the other side will cool off a lot, and as the moon is tidally locked with Earth, such "Moon's days" take a whole full-moon cycle (2 weeks day, 2 weeks night). Such huge difference in "cooking" temperature can cause stress and also contribute to moonquakes. (in both case the Earth is more or less innocent - except indirectly by tidal lock and/or "benefiting" from the moon as a giant meteor shield)
NOTE: I am a Doctor, not an Astrophysicist (in McCoy's voice). So I might be wrong....or I might be lying and covering moonquakes actually caused by nuclear tests done by the Nazis hidden on the dark side of the moon.
Yeah, but "Study shows statistical correlation that might suggest that tidal stress could be among the dozens of other factors that each contribute slightly in favor of stronger earthquakes", though much more accurate, doesn't have the same "oomph" and thus doesn't manage to sell the same amount of eyeballs to the ad-company paying for this fucking article. "The moon will kill with its earthquake power! Cower in terror!! Panic!!! Buy products to feel better!!!!" works so much better...
AMD is pouring resources (documentation, and salary - i.e.: some developers are on AMD's own payroll) on the opensource driver stack.
Nowadays you have 2 solutions:
- the opensource driver which is fully functioning and not so bad on later hardware iteration. For the latest (Polaris) AMD actually managed to pull a fully functioning opensource driver within reasonable time - it was available on launch day! (even not necessarily in most disto - one might need to upgrade to versions packaged on 3rd party repositories). Long term target for AMD is *this* to become the official driver for end-users.
- AMDGPU-Pro: it basically reuses most parts of the opensource driver (e.g.: kernel driver) only the OpenGL library is different (uses the catalyst one instead of Mesa/Gallium3D). (Long terme target for AMD is to only keep this driver around for professionnal users of CAD software that relies on werid stuff) It is MUCH MORE stable that the clusterfuck that early "flgrx" used to be, before the overhaul.
And as *BOTH* rely on the opensource driver which is upstreamed into the mainline kernel, they BOTH work with the latest vanilla kernel. No need to wait for the developer to release a new driver whenever something change in the kernel.
I think Windows and *n*x are both sufficiently protected against simple network sniffing to make that a non-issue.
Unix: depends on configuration. (goes from straight "everybody trust everyone else" like NIS and NFS servers, all the way up to Kerberos - everything is authentified over an encrypted link) (and the home variant: use SSH + keys for everything)
Windows: I've read some very appalling description of how it works. No or not enough encryption.
So I think this adapter does more than just snooping on what comes by - it must change the behavior of the OS in some way.
Accroding to the summary, the key redirects to different (attacker-controlled) name server and Active Domain server (either running inside the USB adapter, or running elsewhere on the network with the key doing redirection of connections) Without proper cryptographic authentication in place, the attacked workstation will blindly trust these servers.
Most typical installtion of Unix services run encryption (e.g.: SSH for remote access, LDAPS for authentication/log-in, even DNSsec is possible for names) or can be authenticated (NFS support kerberos). Such a different server will fail the cryptographic authentication and will be rejected.
Tell me why it should not be done in cellphone battery areas.
In theory, cellphone batteries, because they are made of lithium(*) have EVEN MORE protection than that. Tons of protection directly built into the battery case itself, and in the charging circuit.
In practice: someone somewhere along the supply and/or assembly manage to fuck up enough of these protection.
Thus Samsung joins the hall of shame previously occupied by - Sony (and their incendiary laptop batteries back in the early 2000s) who managed to burn countless laptops, even of 3rd party brand to which Sony was only a battery parts provider (and this fucker contain way much more energy than a smartphone battery. A laptop has the same energy content as a small handgrenade) - and the various no-name Chinese clone maker of self-balance board (the things nick named "Hoverboards") who where focusing so much on making a quick buck that they completely neglected any protection in their rush. (and these fucker pack even more energy than a laptop. THAT's some dangerous fire/explosion hasard).
--- (*) lithium tends to explode if - you overcharge it - you undercharge it then try to charge - overheat - draw too much current while using - put too much current while charging - puncture... Well metaphorically, it will explode if you look it the wrong way.
Some big company is issuing a recall because their gadget can explode. This company is dead serious about it, urging user to keep using their older phone until they receive replacement for the newer-explodey phone.
In short: GADGET CAN EXPLODE, COMPANY WANTS YOU TO REPLACE IT.
You:
A. REPLACE: Hide the thing away in some solid trusted container and bring the shit as fast as you can to the nearest shop for replacement before it explodes.
B. IGNORE: You give it to your kid to muck around - specially because said kid is pre-schooler and has no concepts of "exploding" or "safety".
Any sane person will go for option A.
The only reason that option B doesn't qualify is that you remove the genes from the gene pool by proxy (for DA you need to remove your own genes from the pool before you reproduce. Not remove your offspring after you've reproduced. Aka: don't involve innocent people in the procedure).
Someone should *REALLY* call child protection service over this one. If the (grand-)parents are *THAT* stupid, chances are high that the kid is exposed to other risks due to the irresponsibility and stupidity of close family.
Another alternative is to use proper cryptography between your machine and the necessary server.
I'm not that used to Windows and Active Domain, so I can't comment much.
The Unix equivalent would be to setup LDAPS for the credential validation instead of plain LDAP, with properly signed certificate. The rogue credential server running inside the USB would fail the certificate validation and the worsktation will refuse to use it.
To listen to my existing headphones and charge the phone at the same time, I'd have to buy a Lightning splitter. Neither the earbuds nor the adapter has an extra Lightning port, and they take up the only Lightning port on the device.
Apple missed a HUGE opportunity to bank on the Wireless charging craze. They could have even not needed to go with Qi like everybody else, but invent their own peculiar standard as usual (as Palm did with TouchStone back when Qi wasn't a thing yet). Call-it "iCharge", require an Apple-cryptographically signed NFC chip inside (à la MFI) and bingo, they got a reason for their fan to yet again rebuy all of their accessories (new docks).
Also, I don't understand at all this craze of making smartphone so thin that you can cut cheese with them... the things got screen wider than a paperback book, and users who don't want to break them within 1 month after buying will put them in a cover/case any way.
Keeping smartphone a couple of milimeters thicker would allow not only for extremely useful legacy connectors like the analog audio port, but would even increase space for bigger battery (and add a LED at the bottom of the audio connector so it can double as a digital-out, like some minidisc players used to do).
Slashdot Mirror
The other common "hard" limit is 32GB.
That one is a software limitation.
"SDHC" cards go up to 32GB
"SDXC" start from 64GB
There's no pinout nor SPI difference between the 2.
The only difference is a logical one.
SDHC cards come pre-formatted with FAT32.
SDXC cards come pre-formatted with exFAT, and Microsoft has patented the shit out of it.
So unless the company has paid money to Microsoft, they can't use exFAT and can only advertise "up to 32GB SDHC cards".
But nothing prevents you to buy a 128GB SDXC and :
- either install a FUSE-exFAT driver on your OS if supported.
- or reformat the card with something supported by the OS (depending on the OS: FAT32, F2FS, Ext3/4, BTRFS, UDF, etc.)
So 128/256/512/1024GB will work on most SDHC readers (i.e.: that support more than 4GB plain- SD), but the manufacturer can't advertise it because they lack the patents to the file format that is mandatory to advertise SDXC support.
Oh, you plug it in and it works, but you don't get all the fancy per-game remapping and 30s startup time and 1-2GB RAM consumption.
It works in my case.
The whole stack support them:
- Linux' USB HID driver perfectly supports all the available axis and buttons my the MX Master.
- The SDL library used as a middle ware support multiple extra mouse axis and buttons.
- Quake3 supports binding any commands to any input. (You just get fancy name like "Right Click" on the first few buttons. Then it's more like "M_BTN7"...)
For stupid games that don't have a good re-mapper, there's always things like (rejoystick, joy2key, qjoypad, etc.)
As usual, the crap software that comes with your hardware is crap.
(like the fancy stupid DVD player that came with your optical drive. Simply use VLC)
Most large companies have their very own copy of github in-house. Most commonly these days it's on AWS or some other cloud offering that the company controls for their IT Projects.
I can confirm.
Though we aren't *companies*, most of the universities and research institutes here around (Switzerland) have their own in-house git repository.
Though in our case, a self-hosted copy of * Gitlab * is what is the most popular here around.
And most of the time it's hosted on the universty's/research institute's own server because of complex IP/publishing/secrecy considerations
And if we do it, I can clearly imagin that huge corporation could be doing it too.
And for the record, Google has announced that they've shut down Google-Code for 3rd party project only. Android is *still* officially hosted on their servers.
Their flagship mouse MX Master has quite a few driver issues and also the DarkField sensor seems to have problems with some surfaces.
Drivers? For a *mouse* ?
It's fucking USB HID device (or bluetooth wireless if your laptop has Bluetooth 4.0 LE and you go dongle-less).
It's just a bunch of axis and buttons (though you have to admit that the MX Master mice tend to have a little bit more than the average of them).
You just plug into anything with generic USB HID support and it should work out of the box.
If it doesn't work, you shouldn't blame Logitech, you should blame the retarded operating system that fails at basic USB HID.
The only thing that would require extra software would be:
- Battery status in your status bar. Which is visible thanks to green LEDs on the mouse if self, so I doubt it's that much necessary to have the status-bar widget. /. Maybe you have a script that automatically sends an e-mail to your smartwatch to remind to plug the charger into the mouse when leaving the room)
(Though we're on
- If you want to do some complex re-paring (e.g: pairing the mouse with a Logitech dongle that wasn't initially designed for it). /. and this might actually be the case. I would point you in the general direction of Solaar)
(Again, this is
- Circumventing a piece of software that has asinine key/button binding. (e.g.: that has some stupid arbitrary hardcoded maximum of 3 mouse button, thus preventing you from binding the extra butons to additional functions). But then, any keyboard/mouse/joystick button remapping software would be plenty enough.
Logitech Quality has declined measurably recently, that's my experience anyway.
I have a strong impression that Logitech doesn't the same build quality in all markets.
(I've noticed difference in products between US market, and Logitech's home Switzerland)
Could anyone else confirm ?
Meanwhile, http://uj3wazyk5u4hnvtk.onion/ continues to work (as long as your have a Tor proxy running, obviously)
I mean is doing.
Their next iteration (Pebble 2 / Pebble Time 2) is *still* only bluetooth connected and with limited low-power processing.
But if you want to go smartphone-less, instead of trying to pack extra smartphone functionality into the watch itself (3G connectivity, Wifi support, whatever...)
they also develop the Pebble Core a small screen-less android powered wearable computer.
Essentially, the screen-less iPhone of your joke, but for real, and powered by Android instead of iOS (and not called the "iPhone Shuffle").
Yes, we should eliminate all those things which are working great, so there are no ports, because... reasons
He just gave you a reason, smartass: so you can bring the phone underwater.
I have quite a few gadgets that can go underwater (not just splash resistant, but water resistant: full IPX7-certified immersion capable) that do feature ports.
Including the e-Book I gave present to my mother (Kobo H20) : that has ports - a microUSB charging/sync port, and a microSD card port.
(As long as the rubber cover is over them, and isn't currently open to swap the microSD card, this thing can survive 1m under water for half an hour)
That also includes my speaker (UE Megaboom): it has a microUSB and a 3mm Audio stereo analog jack and a huge honking battery rated at a whooping 200Wh.
Everything that Apple deems necessary to remove just to make their phone splash resistant.
And this speaker is not only resist spalsh but can be forgotten for some time at the bottom of my above-ground pool. And will keep playing music (though not very audible due to difficulty of transferring sound air->water->air again).
I gave also a photo camera as a present to my girl (Olympus TG-4) this thing is rated at 15m depth (you could go diving with it). And it still has a USB connector, a SD card port, and Audio/Video connection (though the port is digital : it's a micro HDMI, not analog jack port) and fast and easy end-user replaceable battery. Though for the record it use plastic doors instead of rubber covers to protect the ports.
So again, can you explain me why Apple needs to remove functionality that is used everywhere (just go to any student party and watch how often people plug their smartphones with their music playlists into the analog jack of whatever speaker system is laying around*) just in the name of making the damn thing splash resistant ?
Oh yeah, I know: their obsession of making the phone thin enough to slice cheese. A rubber port cover would stay in the way.
(It's the same company that obsesses with ridiculously small SIM card formats, even if the whole device isn't fitting directly in the ear - Lt.Uhura's style)
---
(*) Apple removing the audio jack from their hardware would be like removing the floppy drive from their "canddy" iMacs... before USB flash drive got available.
Currently only some portable speaker system feature bluetooth. Even less are hassle free (NFC-based "touch to connect").
There are tons of old speaker systems in student dorms only accepting analog inputs. (Or even more ironically: accepting audio from the old (non-lightning) Apple connector that isn't produced anymore).
Luckily Bluetooth to 3mm jack / -to stereo RCA (and most ironic: to Apple 30-pin) audio dongle are cheap from China over ebay, so eventually this situation might get more jack-less friendly.
Sony claims prior art in fire-inducing lithium batteries.
(And "hoverboard"-style self-balancing rollwe's opinion?
"Under 100kWh. Meh. Amateurs")
This is the sound piece in question >> It's called "plim plim" in Brazil.
And that sounds very close to one of the alternative chimes available on PalmOS devices by Palm Inc. (was it the "SciFi" chime ? I'm not sure anymore),
available at least since the early 2000s (on my Palm IIIc back then).
Probably other PDAs have use a similar chime even earlier (could some e.g. Psion / EPOC user confirm ?)
Though trademark law doesn't have a concept of "prior art", it does have something similar to "defend or lose it".
Currently this sound isn't distinctly associated with their branding (at least outisde of Brazil. Maybe there this "plim-plim" is immediately associated with the brand and is never heard anywhere else). It's not a jingle that is specific to grupo globo, it's something that has been heard around countless times.
It's completely generic and the EU is completely right at laughing them out of the court.
I got 2 people in my group of friends (which isn't that big, those 2 are pretty much my closer circle of friends) who are anything but geeks (one a mechanic, one an artist) who BEG me to show them how to get away from Windows 'cause they don't want to deal with Win10 and all they do with their computers is pretty much playing games (and watching porn, of course). Now, doing the latter isn't the problem, but I can NOT at the current moment recommend to a non-technical person playing his games in Linux.
Current best bet are:
- follow closely what Valve is doing (both in term of software: Steam, and SteamOS - and in term of hardware: Steam controller)
- stick for now to a distro that is immensely popular: Ubuntu would be a somewhat good bet for now.
- stick to hardware that is immensly popular, and check online forums, etc.
- AMD since the Polaris generation is safe for out-of-box.
- check games that use an engine that is well ported on Linux (something popular like Unity / Unreal / Cryengine / older idTech / etc.) rather than a game that use a homegrown engine that is only rushed to linux, or use actually a bad-quality wrapper.
- Phoronix is a good place to get info about hardware, etc. their forums is also a possible place to ask around for help.
(continuing split)
support for 3D audio is barely there
The only manufacturer still making hardware with 3D Audio is Creative and they do put some half baked broken OpenAL drivers that might work when it doesn't crash everything (thanks for the effort, Creative... next time just release the docs to dev and throw some salary)
Nearly everything else is actually plain simple DAC hardware that is 100% handled by Asla and pulse (because there's no actual 3D hardware or anything. it just converts a 44100 or 48000 bitstream into analog signal).
All the "3D Audio" is actually entirely done in software, often using an official service (I think DirectSound is the official one in Windows ?) and linux *does have* an OpenAL implementation, or much more likely entirely done in the game's engine.
Currently, linux dev have done all the possible.
It's now up to:
- Creative to fix their shit or release XFi docs and pay a kernel developer.
- Game developer to try to leverage what is available (generally; either use OpenAL, or have the game engine do the processing and output to SDL which will handle the back-end interaction ALSA/Pulse)
and support for any kind of gaming hardware, from macro keyboard and multi-button mice to racing wheels to flightsticks to VR-hardware, is practically non-existing.
Razr is releasing docs (and I think giving donations to developers ?) so there's support underway, for all features including programmable coloured lights.
(But this has started only recently, so I have no idea, how far is the out-of-box experience).
Logitech is sometimes donating hardware to developers and thus lots of their hardware is working, including advanced features like force-feedback. But it depends on models, so sadly you currently also need to track down which exact model before buying.
Most of the hardware is generally still working through generic HID drivers that will expose simply a collection of buttons/axis, so they are still somewhat usable even without the advanced features (force feed-back, special macros).
But than the problem is again the game developers: they lack specially in the button assignment department, most games can't even assign buttons correctly and refuse to work with anything that isn't an Xbox controller.
And I tend to use a pretty simple setup when I play (regular mouse + keyboard, or regular joystick) so I don't pay attention to all the "macro" craze and thus can't give much more informations.
My general recommendation would be, in order:
- use something actually developed with Linux in mind and with strong software support, e.g.: Steam controller
- use simple vanilla controllers (keyboard + mouse, or joystick/wheels that work with basic HID).
- use something with active development: Razr and check hardware support (it has only started recently)
- use something with active development: Logitech and check hardware support (not everything is supported)
- use something widely popular like an Xbox controller: Microsoft hasn't moved an inch, but the thing is so popular that there is support (just check).
Look at what walled-off, closed atrocity Apple has made out of Unix, and the people buy it. Why? 'cause it works.
Fixed that for you.
Apple also runs an Unix, but it is based on BSD kernel (running atop of a Mach micro-kernel) instead of Linux. Very probably for licensing reasons (BSD vs, GPL).
Using CSS drivers on an equally proprietary piece of hardware is a VERY small price to pay to get more people to join the ranks of Linux users.
The problem is that you need *to be able* to run said CSS drivers. And that's where things get problematic.
You see, Linux is a different beast than Windows where people still run the same Windows XP 15 years later.
Linux kernel get improved/fixed/modified/modernized. Some API/ABI change.
For opensource drivers, that are part of the kernel source tree, and that get dedicated developer to maintain (like those on AMD's payroll), that means that the driver will get also upgraded along and they will work with whatever latest vanilla kernel you get.
For closed source drivers, that means that they get regularly broken, and the hardware maker that wrote the driver need to get their shit done *ON TIME* so the closed source driver also works with Linux Kernel 4.8
And that's the problem here.
Nvidia closed source work atop of a shim that they write themselves and that needs to be modified to work with every single last kernel version out there.
So currently your only solution is to run Ubuntu to which Nvidia happen to deliver preview version of drivers matching the kernel version of the distro.
For everything else, the situation sucks, and "nouveau" is the only remaining opensource driver that will be more or less updated and will at least load on any kernel version.
Contrast the situation with AMD:
Since at least GCN 1.2 (previous generation of Radeon hardware - also doable with GCN 1.0 and 1.1, but currently not already mainstream), all drivers stack run atop the same kernal driver.
This driver is opensource.
This driver is in the main kernel tree and thus is available in any vanilla kernel.
This drivers gets maintainers/developers some even paid by AMD themselves.
Thus even on the latest kernel (4.8) you get a Radeon kernel module (amdgpu.ko) that actually works.
With the current generration of card (Polaris) on launch day, you could use it out of the box on your latest distro, it works.
- you either use the stock opensource stack, that uses Mesa/Gallium3D to provide OpenGL 4.3 (and in the case of Polaris, that has performance close to the closed source offer)
- or you swap the dynamic library providing openGL with AMDGPU-Pro, the official closed source openGL which runs atop *the same kernel driver, so also out-of-the box on the latest kernel version*, which will provide OpenGL 4.5 (and all the weirdness necessary for CAD software).
This is something you can buy and use *right now*.
Intel, although not playing in the same field, *also* provides launch day opensource support in upstream kernel.
Why couldn't Nvidia bother to move their ass and do the same? Oh way, that would require them to develop an actual linux driver, instead of simply recompiling their Windows code on linux and use a shim to pass along whatever is needed. (And thus regularly breaking/missing some corner case feature that aren't exactly handled the same way on linux... cough.. Optimus! ...cough...)
Because more people using Linux means more reasons for hardware makers, DESKTOP HARDWARE makers, to put out sensible support for their hardware.
But Linux developers currently don't have the resources to do the hardware makers' job.
If there's no launch day driver available, because closed-source code need to get upgraded and/or because there are no developer save a few volunteers,
nouveau is all you could get.
- Either you put up with nouveau until Nvidia does their shit and then you do
they fire the knowledge last year
Well, at least, some of the earliest casualties in the "Knowledge" part - those who worked on the Linux effort with Maemo/Meego - have since then founded Jolla and had some relative success in making a good full-blown GNU/Linux OS for phones with Sailfish OS.
(Happy user of a Jolla1 phone, looking forward for an official port to Fairphone2 including the Android Apps support bit).
The problem is that the kernel is always progressing.
So for the GeForce binary blob to work, you need to have an upgrade that works with the kernel version provided to you by your distro, which very often IS NOT available at the time of release.
Since Nvidia have discontinued their own 2D only "nv" driver, "nouveau" is the next best thing that is available in a modern kernel out-of-the box.
But that thing is developed by volunteers only, with nearly no help from Nvidia themselves.
So you're basically stuck to using only offical builds of Ubuntu, because at least Nvidia provide preview version of drivers in a timely manner.
Compare the situation with AMD, where Mesa/Gallium3D *AND* the closed source AMDGPU-Pro libraries both use the same opensource kernel module, that is upstreamed (its source is in the official kernel tree) and whose developpers include people on AMD's payroll:
- currently, you know that modern cards that you buy now (e.g.: the current Polaris) work out of the box with any modern vanilla kernel.
- drivers just boils down to swapping with dynamic library will provide OpenGL. Current Mesa 12.1 provides you with OpenGL 4.3 out of the box, or you could swap for the AMD's closed source AMDGPU-Pro library (which sits attop the SAME kernel driver which, as said, works out of the box) which will provide you the current latest OpenGL (4.5) and include all the weird quirks that you might need if you're working with CAD software.
- forgot to mention: on official launch day, performance of both Mesa and AMDGPU-Pro libs were reported to be relatively close.
That's the kind of dedication I'm willing to pay for.
waiting on *soon* promises
Polaris is out there now.
Mesa 12.1 works with it right out of the box at OpenGL 4.3 level.
So unless you're on some ancient distro (e.g.: you're locked on some LTS cycle for version-stability reasons) you can have it right now.
(otherways you wait until *your distro upgrade*)
I'll just stay with what always just works for me under Linux: Nvidia.
You mean this piece of shit in my laptop which was a nightmare to upgrade with the official drivers (always needing to wait until Nvidia finally release an upgrade to the kernel driver to follow evolution of the upstream kernel) with very slow pace (my laptop isn't covered by the latest drivers anymore, I need to wait for the older legacy driver to get updated too) and that on occasion has massive graphical corruptions ? (And that's when the text console doesn't outright stop working ?)
And thus is completely unusable on any rolling-release distro (e.g.: Opensuse Tumbleweed ?)
Whose opensource drivers are developed by a bunch of volunteer to which Nvidia gives no documentation nor any other help except once every blue moon when they throw a bone with some Tegra-related docs that also happen to miraculously help the GeForce efforts too ?
(And thus doesn't handle the blanking of my laptop ?)
Sorry, but no thanks.
Nvidia is unusable on anything that isn't :
- a standard desktop with a generic GeForce card
- or evetually a very recent laptop with a very classical build
- running the latest official Ubuntu.
Meanwhile Radeon card gets docs and salaries by AMD.
Latest Polaris is a smooth "out-of-the-box" experience.
And in the past, the opensource drive has always been a smooth experience on the various build I've used it (though, I need to conceed, good OpenGL levels (e.g.: 4.x) have only been achieved in the recent iterations of Mesa).
I can easily imagine a lot of affected Note 7 owners -- approximately 2.5 million of them -- weren't exposed to the recall message for some reason.
We're speaking about smartphones.
- As per the various laws that were put into action in the name of "protection against terrorism" after 9-11, lots of countries (including the US) require tracking off all users of cell phone services (even pay as you go) to be properly registered. That means, at some point during purchase, this smartphone was associated with a SIM card that has an identity linked to it.
It's not necessarily the SIM card that finally ended up in the phone (e.g.: the grand parents might have bought it, but kept their old phone with the SIM in it, and might have given the smartphone as a present to the 6yo kid), but easy to track down that the grandparents were present at the moment when the phone was bought.
Through this way it is possible that the store chain that sold the phone (or the company that is keeping track of the cell phone customer for them) could track every last buyer of the phone and tell them. (They won't be able to tell the kid, but at least they can reach the grand parents who bought it).
If the way SIM-card holder registry is done in the US can't be used to track down users in case of emergency, then it's really a piece of shit that can't even stand up for the reason it was invoked (though I strongly suspect that this is the case. Just another privacy invading policy that can't be actually used against terrorism or for emergencies).
- Due to the way radio-wave licensing works, there are part of the phone radio circuits (parts of the firmware running in the modem, and the firmware running in the SIM card) that aren't update by Google / Samsung (they aren't part of the Android installation installed on the smartphone). ...cough...)
They are not supposed to be end-users accessible.
Only holder of license for the radio channel can access it. It's the carrier's job to send over-the-air updates to this part of the modem firmware, and to the SIM card. (When not straigh sending a new physical SIM card, because after 10 years the previous one is too old and useless to run anything compatible... been there...) (Also hope that your phone's SoC doesn't use it's modem chip *as a Northbridge/main controller* and the modem-firmware update causes freezes/crashes in the OS... cough... Qualcom!
Thus at any time every cell tower knows exactly which phone (and exactly which *hardware model* of phone) is connected to it (and might send such OTA updates).
Cell tower can also send not only SMS messages, but direct service message (a full screen text/warning that bypasses the SMS system).
Also keep in mind that, even without a SIM card, a smartphone *will* establish contact to a cell tower (at least to provide emergency calling services).
Thus it's trivially possible (with a bit of cooperation between carriers) that a message is delivered straigh to the screen of all concerned smartphones (i.e.: a specific IMEI range, or a certain firmware revision...)
Even if the 6yo can't read the "WARNING: this phone is about to explode" message, he'll probably complain to the nearest adult that the pop-up is disturbing his session of Pokémon.
As a comparison, 1 year ago, here around a couple of store chains needed to recall quick-build above-ground inflatable spas/whirlpool due to potential electric shock hasard in the heating system.
It's not even an IoT connected thing, there's no automatic way to remotely contact users / display warning on the console. It was entirely done through public communication / announcement.
Still, the whole defective batch was successfully recalled without any causality.
But it might be that germanic speaking countries might be a little bit better organised than the US.
Still, if they can manage to recall a dumb spa, whereas Samsung and the US cell carriers are unable to recall a connected device that can remotely display message and whose buyers are supposed to be easily traceable (for terrorism/emergency reasons), then there's a deep problem somewhere.
There's a lot less likeliness of Moonquales happening:
- The Moon, as mentioned by others is tidally locked with Earth.
i.e.: It's always the same side facing Earth
that means that Earth tidal stress is always "pulling" the exact same part.
thus no change in this "pull" and therefore no directly-earth-caused tides on the moon.
- The Moon is a huge solid rock, it doesn't move much, whereas the Earth is pieces of solid crust all covering a molten mantle on which they more or less float (the "more of less" part being when they bump into each other, rub against each other, or one dives under another).
Thus Earth has a tectonic seismology (the tectonic plate doing their above mentioned shit instead of peacefully floating next to each other), on which a whole bunch of factors (including in parts the tidal stress exerted by the Moon as maybe suggested by this study) can act as trigger to release to stored shear stress (causing Earthquakes).
Whereas the Moon, by being a solid block of rocks can't do that.
- On the other hand, the Moon has no atmosphere :
Any falling rock will smash straigh on the surface. Such meteorite contribute heavily on the moonquakes.
The part of the moon that face the Sun will be backed in light, while the other side will cool off a lot, and as the moon is tidally locked with Earth, such "Moon's days" take a whole full-moon cycle (2 weeks day, 2 weeks night). Such huge difference in "cooking" temperature can cause stress and also contribute to moonquakes.
(in both case the Earth is more or less innocent - except indirectly by tidal lock and/or "benefiting" from the moon as a giant meteor shield)
NOTE: I am a Doctor, not an Astrophysicist (in McCoy's voice). ...or I might be lying and covering moonquakes actually caused by nuclear tests done by the Nazis hidden on the dark side of the moon.
So I might be wrong.
Yeah, but
"Study shows statistical correlation that might suggest that tidal stress could be among the dozens of other factors that each contribute slightly in favor of stronger earthquakes", though much more accurate, doesn't have the same "oomph" and thus doesn't manage to sell the same amount of eyeballs to the ad-company paying for this fucking article.
"The moon will kill with its earthquake power! Cower in terror!! Panic!!! Buy products to feel better!!!!" works so much better...
Maybe you should upgrade your drivers:
AMD is pouring resources (documentation, and salary - i.e.: some developers are on AMD's own payroll) on the opensource driver stack.
Nowadays you have 2 solutions:
- the opensource driver which is fully functioning and not so bad on later hardware iteration. For the latest (Polaris) AMD actually managed to pull a fully functioning opensource driver within reasonable time - it was available on launch day! (even not necessarily in most disto - one might need to upgrade to versions packaged on 3rd party repositories).
Long term target for AMD is *this* to become the official driver for end-users.
- AMDGPU-Pro: it basically reuses most parts of the opensource driver (e.g.: kernel driver) only the OpenGL library is different (uses the catalyst one instead of Mesa/Gallium3D).
(Long terme target for AMD is to only keep this driver around for professionnal users of CAD software that relies on werid stuff)
It is MUCH MORE stable that the clusterfuck that early "flgrx" used to be, before the overhaul.
And as *BOTH* rely on the opensource driver which is upstreamed into the mainline kernel, they BOTH work with the latest vanilla kernel. No need to wait for the developer to release a new driver whenever something change in the kernel.
I think Windows and *n*x are both sufficiently protected against simple network sniffing to make that a non-issue.
Unix: depends on configuration.
(goes from straight "everybody trust everyone else" like NIS and NFS servers, all the way up to Kerberos - everything is authentified over an encrypted link)
(and the home variant: use SSH + keys for everything)
Windows:
I've read some very appalling description of how it works.
No or not enough encryption.
So I think this adapter does more than just snooping on what comes by - it must change the behavior of the OS in some way.
Accroding to the summary, the key redirects to different (attacker-controlled) name server and Active Domain server (either running inside the USB adapter, or running elsewhere on the network with the key doing redirection of connections)
Without proper cryptographic authentication in place, the attacked workstation will blindly trust these servers.
Most typical installtion of Unix services run encryption (e.g.: SSH for remote access, LDAPS for authentication/log-in, even DNSsec is possible for names) or can be authenticated (NFS support kerberos). Such a different server will fail the cryptographic authentication and will be rejected.
Tell me why it should not be done in cellphone battery areas.
In theory, cellphone batteries, because they are made of lithium(*) have EVEN MORE protection than that.
Tons of protection directly built into the battery case itself, and in the charging circuit.
In practice: someone somewhere along the supply and/or assembly manage to fuck up enough of these protection.
Thus Samsung joins the hall of shame previously occupied by
- Sony (and their incendiary laptop batteries back in the early 2000s) who managed to burn countless laptops, even of 3rd party brand to which Sony was only a battery parts provider (and this fucker contain way much more energy than a smartphone battery. A laptop has the same energy content as a small handgrenade)
- and the various no-name Chinese clone maker of self-balance board (the things nick named "Hoverboards") who where focusing so much on making a quick buck that they completely neglected any protection in their rush. (and these fucker pack even more energy than a laptop. THAT's some dangerous fire/explosion hasard).
--- ...
(*) lithium tends to explode if
- you overcharge it
- you undercharge it then try to charge
- overheat
- draw too much current while using
- put too much current while charging
- puncture
Well metaphorically, it will explode if you look it the wrong way.
Some big company is issuing a recall because their gadget can explode. This company is dead serious about it, urging user to keep using their older phone until they receive replacement for the newer-explodey phone.
In short: GADGET CAN EXPLODE, COMPANY WANTS YOU TO REPLACE IT.
You:
A. REPLACE: Hide the thing away in some solid trusted container and bring the shit as fast as you can to the nearest shop for replacement before it explodes.
B. IGNORE: You give it to your kid to muck around - specially because said kid is pre-schooler and has no concepts of "exploding" or "safety".
Any sane person will go for option A.
The only reason that option B doesn't qualify is that you remove the genes from the gene pool by proxy (for DA you need to remove your own genes from the pool before you reproduce. Not remove your offspring after you've reproduced. Aka: don't involve innocent people in the procedure).
Someone should *REALLY* call child protection service over this one. If the (grand-)parents are *THAT* stupid, chances are high that the kid is exposed to other risks due to the irresponsibility and stupidity of close family.
Another alternative is to use proper cryptography between your machine and the necessary server.
I'm not that used to Windows and Active Domain, so I can't comment much.
The Unix equivalent would be to setup LDAPS for the credential validation instead of plain LDAP, with properly signed certificate.
The rogue credential server running inside the USB would fail the certificate validation and the worsktation will refuse to use it.
To listen to my existing headphones and charge the phone at the same time, I'd have to buy a Lightning splitter. Neither the earbuds nor the adapter has an extra Lightning port, and they take up the only Lightning port on the device.
Apple missed a HUGE opportunity to bank on the Wireless charging craze.
They could have even not needed to go with Qi like everybody else, but invent their own peculiar standard as usual (as Palm did with TouchStone back when Qi wasn't a thing yet).
Call-it "iCharge", require an Apple-cryptographically signed NFC chip inside (à la MFI) and bingo, they got a reason for their fan to yet again rebuy all of their accessories (new docks).
Also, I don't understand at all this craze of making smartphone so thin that you can cut cheese with them... the things got screen wider than a paperback book, and users who don't want to break them within 1 month after buying will put them in a cover/case any way.
Keeping smartphone a couple of milimeters thicker would allow not only for extremely useful legacy connectors like the analog audio port, but would even increase space for bigger battery (and add a LED at the bottom of the audio connector so it can double as a digital-out, like some minidisc players used to do).
Google owns Widevine, not Adobe. Did you mean Primetime?
Ooops. My mistake. I confused the DRM plug-ins.
Thank you for rectifying
(Tells you how often I use DRM in my day-to-day life)