Forgot your password?
typodupeerror

Comment: Re:Choosing Sides (Score 1) 758

by Anaerin (#47751655) Attached to: Choose Your Side On the Linux Divide

1. ok, so it needs a bit of rework to multithread its process-starting system. I that significantly more difficult that rewriting the entire loader?

It needs more than just "a bit of rework", it requires a complete overhaul. Which is what SystemD is.

2. So it needs an extension to monitor services. Technically, I think this is better handled by a different task, one that is more into monitoring rather than blindly just continually-restarting a service that's crashed due to some external dependancy failure. Again, its not much of a task to add this than it is to rewrite the entire loader.

So, instead of one system handling bringing up a service, you want two. One to bring it up at runtime, and another to bring it back up if/when it crashes.

What SystemD is doing is, rather than using a stupid-simple for loop to start things, it's using an event-based system. This also means you can have things happen like "Network comes up, start network-dependent services", "USB printer plugged in, start print daemon", "Device hot-plugged, start dependent services". These are all things that are missing from the current SysV system, and require extra tools, along with their associated extra configuration and extra points of failure.

Comment: Re:Choosing Sides (Score 4, Insightful) 758

by Anaerin (#47751173) Attached to: Choose Your Side On the Linux Divide

You're posting a lot on the drawbacks of SystemD, but you're not including any of the drawbacks for SystemV.

  • SysV starts every task and service one at a time, waiting for the previous one to finish before it starts the next. This is fine for single-core single-threaded workloads, but most systems these days are multi-core. It also means that startup is slow. SystemD, on the other hand, can (and does) start up services in parallel, making sure that dependencies are resolved before starting the next item. This makes SystemD MUCH faster at booting.
  • SysV only handles the initial bringing up, and the final tearing down, of services. Any service failures are not noticed, any errors are not dealt with. SystemD monitors services even while they're running, and can re-start them if they fail, handle crash reporting and log rotation. This makes SystemD more error resilient and stable.
  • SysV only runs the services, and doesn't care about their configuration, so basic configuration and dependencies have to be handled in a myriad of ways, different for every service. SystemD attempts to be a central repository for ports (and handling the necessary firewall rules for them), configuration files, logging locations, and the rest. While this does make startup a touch more difficult to configure, it makes the system as a whole a lot simpler to deal with.

Comment: Re:Most Tv's can already do that by themselves (Score 2) 112

by Anaerin (#47659745) Attached to: Xbox One Will Play Media from USB Devices, DLNA Servers
It doesn't have to be a Samsung app (Which, by the way, is kinda awful). DLNA means it can be any DLNA server, like Serviio (for instance) that also happens to have the wonderful feature that it will auto-transcode into a suitable format if your TV/Console/Whatever doesn't support it. Which means the PS3 and X360 (Also supported by Serviio) can also get in on this party.

Comment: Re:110 or 240v (Score 1) 260

by Anaerin (#47511681) Attached to: Google Offers a Million Bucks For a Better Inverter

Yes, because the US cheats and uses 220 split-phase to provide 110 power. Most everywhere else that needs high power uses 3-phase, as it's smoother, easier to produce and rectify, and just as safe to transmit.

3 phase makes electric motors more efficient, and that's it. Technically, you could have as many phases as you could imagine having... each making the motor a tad more efficient. But they are not "smoother" and don't improve transmission.

3-Phase AC produces a smoother (considerably less ripple) DC current pattern when rectified than single or split-phase AC.

Comment: Re:A500+, A600, A1200 (Score 1) 192

by Anaerin (#47498381) Attached to: The Almost Forgotten Story of the Amiga 2000
And if you can still find them, the A1200 has expansion options up the wazoo, including RTG graphics, 16-bit soundcards, Ethernet adapters, dual-channel PIO Mode 4 IDE interfaces (To augment/replace the single-channel on-board IDE), Keyboard converters, even full PCI Backplanes or Zorro II/Zorro III/ISA/PCI backplanes which support many modern graphics cards, among other things.

Comment: Re:SCSI madness (Score 1) 192

by Anaerin (#47498269) Attached to: The Almost Forgotten Story of the Amiga 2000
The thing is, that was SCSI's fault, not the Amiga's. Dealing with DIP switches and termination was a problem no matter what OS you ran, or who made your hardware. Amigas had AutoConfig, which would (for Zorro I/II/III cards) allocate DMA, Interrupts and IRQs automatically at boot-time, along with auto-loading firmware and drivers. It was so good, in fact, that Intel copied it and called it "Plug 'n Play".

Comment: Re:Focus? (Score 4, Interesting) 42

Well, as the phone (that you slot into the "front" of the contraption) already has a suite of magnetometers, gyroscopes and accelerometers, it can quickly and reasonably accurately track head movement. There's also a hold for the phone's camera to look out, so it can be used for AR purposes (and the camera can be used to augment the head-tracking above).

Of course, what I'm waiting for is for someone to add some kind of "Steam In-Home Streaming" (or something similar) support for the phone, so you can run Oculus-supporting games on your PC and have the video sent over WiFi to your phone/HMD, which would be the perfect "Killer app" for this system. And the chances are, someone (Possibly even Valve) are working on this right now. It might exacerbate the lag issues (what with WiFi latencies and encoding/decoding latencies being added to processing latency), but it would be palpably better from a "freedom of movement" standpoint (and possibly a weight standpoint) than a Rift.

Comment: Re:Hey, AmigaOS has a gift for you... (Score 1) 126

by Anaerin (#47330807) Attached to: Google Demos Modular Phone That (Almost) Actually Works

Okay, so here's a scenario for you. I've just built a nice new Ara phone. It has a computing module, a camera module, an LTE/GSM+SIM module, a 802.11a/b/g/n/ac module, 128GBs of storage, a touchscreen and a fingerprint reader.

It's the first time I've put this device together, with brand new parts out of the box. How am I meant to download the drivers? I can't use the WiFi, or the cellular modem, I don't have drivers for them yet. And I can't display any kind of configuration, because the display isn't set up either.

The "phone" has the main OS pre-loaded (I'm presuming a bare kernel on the Computing module, as that's what would decide what version of binaries and the like would be needed), so it can boot, but there's no functionality on it to mount the storage, or bring up the display, or even to start the WiFi and/or Cellular data, because there's no drivers for that yet.

The way I'm suggesting means that all the drivers are compartmentalized and available from first boot. The moment you slot a new module in it's ready to use. And while you can update the software onboard, you don't need to download software to get the system up and running. The on-module flash would also be locked into read-only mode for regular operations, and only unlocked as read/write when an update is required.

Comment: Hey, AmigaOS has a gift for you... (Score 1) 126

by Anaerin (#47329737) Attached to: Google Demos Modular Phone That (Almost) Actually Works

It's called Autoconfig. Essentially, Autoconfig does IRQ and address assignment (And is so good at it that Intel copied it for their "Plug'n'Play" system), but Autoconfig does more than that. It also initializes the firmware and loads it in. And the firmware for each device then contains the necessary libraries to (at the very least) get the hardware running. So, for example, hard-drive controllers get their drivers loaded and the hard-drive becomes available as a boot device, network cards are initialized enough that PXE booting works, graphics cards are started up and displays initialized, sound cards are initialized, etc.

So, essentially, every piece of hardware would take care of it's own drivers and NVRam config. WiFi module has WiFi drivers on it. And so on. The configuration software was not included on-device, but there's next-to-no reason why that couldn't be on-device too, as the price of flash is extremely cheap these days.

Comment: Hard enclosure? (Score 1) 86

by Anaerin (#47133023) Attached to: A Bike Taillight that Goes Beyond Mere Taillighting (Video)

Why is he proposing a hard enclosure for this? It would be a lot more practical (not to mention more convenient and lighter) to sew these into the covering flap of a messenger bag or similar (like he had), using easily-obtainable RGB LEDs on flexible PCB strips (like he was using) that are already sealed and watertight. The only difficulty would be sealing the connecting points, but that wouldn't be too much of an issue, and if you're going whole hog and making from the SMT parts up (instead of repurposing already made components) you could make a square sheet of flexible PCB for the whole thing.

Also, aren't there laws against putting flashing/strobing/colour-changing signs and lights in front of people's faces while they drive? While better visibility is good, techno-disco-light-shows distracting all and sundry on the road is bad.

Remember the good old days, when CPU was singular?

Working...