Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[BigD145]

It did ask the first time, but ever since it will put these empty folders at the place I told him, and doesn't ask anymore

I didn't know nvidia made garbage folders like that at all. I can't remember them ever doing that. My search turned up nothing.

[dragdeler]

hmm... I couldn't tell you if they're allways named the same and allways entirely empty, but I 100% associate these folders with updating gpu drivers since like late 2018 when I got this machine.

This is part and parcel with that "The user is too stupid to do the needful, so we will do everything for them-- Isn't that GREAT!?" trend.

They're training a generation of technologically illiterate people, I'm all for making things easier to use, but the iconification of everything goes so far beyond as to probably having an impact on alphabetism alltogether. I allways rant about this stuff but yeah, one day most people wont even think of things as files, not to mention know where they are. Go check teachers ranting about homeschooling, here is a quote I enjoyed: "Just because mom wouldn't even dare go on the store to download an app doesn't mean you're technologically competent". Ironically though among the list of complaints was students being unable to create logins containing the right info and syntaxe - instead of justing handing them out readymade  . In conclusion we are doomed.

[feelotraveller]

I allways rant about this stuff but yeah, one day most people wont even think of things as files, not to mention know where they are.

One 'word' for you "regedit". (Actually two words, but I digress...) Oh, and a bucket of salt, hopefully directly on the wound. Only confirmation here is that windoze is smelly garbage. Please eject promptly.

[Duuvian]

I'm looking for a typing keyboard that has a variable keypress. Is there such a thing? I'm not sure what the exact term for variable keypress is, but by that I mean like how on a piano you can do half keypresses and for less sound.

Also I imiagine it could be useful in driving and flying video games where on/off switches make controls jerky with keyboard sometimes because you are stuck at one angle of turn unless you tap or it's one of those hellish toggle step control schemas like the rudder in the old Red Baron 2.

I noted that playing my typing keyboard in FL Studio the guitar either blares a note at whatever settings it's at or it's not playing a note, with no ability to have quieter (or other effects variable by amount of key depression) notes from lighter keypresses. Since it's a synth guitar, it should be capable of increasing and decreasing volume or whatever in a way without a new strike of the strings like a real guitar, which would be simulated by a new keypress. I could set that up by programming (via graph music is fun) the right knobs to turn on various UI as I play but that is a real pain in the butt to do free form. Maybe guitar was a bad example... Anyways though it would be nice to be able to throw in quieter notes for more variety as well as to be able to "wiggle" a key so the same note receives whatever modifier key depression is tied to, if that level of wizardry is even possible in FL Studio.

If not, it might be useful for games maybe? Surely not a boondoggle of a really niche keyboard that doesn't exist?

[dragdeler]

I would be hell of surprised if there exists such a thing



Tssss


Afaik the original xbox had pressure sensitive buttons on the right side, and I don't think they discontinued them for cost reasons... buttons are kinda supposed to be 0-1.


Alternatively you find yourself a midi controller or pianokeyboard. With bome's midi controller you can map keyboard keys "globally" to it, not just in single programs.... But be it the readymade keyboard from the link or some hacked together stuff... I think there won't be a lot of games that play along if you map analog to a keyfunction, idk I never tried remapping an analog key from my gamepad to to a normal key function in a game, but I'd reckon 98% of things still would only work as vulgar key.

[Mephisto]

In the strictest sense, tactile switches kind of have a half-press. It doesn't do anything extra, just allows for quicker typing.

Check out some animations to see how linear and tactile differ.

[wierd]

I am at a loss for how you would even have the computer interpret a keypress with a soft press.


About the only thing that really makes sense, is to have it somehow change the typematic rate, or maybe, if you POUND on the keys, have the computer interpret it as bold font?


Personally, I would say this is something that calls for a custom electronics project, with say, an ATMega or something driving it.  Have it present itself as a HID class USB game pad, with 103 trigger switches.

[Starver]

"Fuzzy typing".

As you type, with varying amounts of pressures detecected [1] it can accept 'possible' and reject 'unlikely' keys that are merely brushed and save the user the need to waste time fully depressing everything.

See also the Ig Nobel Prize for Computer Science, 2000!  And of course there's an xkcd or two that might be considered relevent.


(In reality, I really would like a keyboard to be definite. While at the same time I have pressed down so much on more than one set of cursor keys in (e.g.) driving games where I really want more speed/turning/etc that I've actually worn the printed arrows out, warn the textured surface flat and even worn through the surface to reveal part of the void within...)


[1] Which can also be used to somewhat identify the operator. Already there are systems that profile the typing cadence and transitions (c.f. a telegraph operator's "fist"). No longer would it be possible for the (pro/an)tagonist in a film to dust the password entry system[2] and magically come up with what the legitimate user would have entered for their own successful

[2] Personally, when it comes to the fancy electronic-screen style fancy keypads, like the one in Entrapment (Connery/Zeta-Jones, 1999) with the characters necessary for "Don't use a cannon to kill mosquitoes", I'd have thought adding a randomised position (perhaps additional words, perhaps multiple key-phrases (per authorised person (+'duress' phrases to activate silent alarms?)) with the current choices available dictating which is applicable to this instance.

[Duuvian]

Tssss

https://store.coolermaster.com/us/mk850-red
M4: Advanced Flight Mode - Aimpad is enabled in all 8 keys: QWER, ASDF. WASD will function like the left analog stick, Q and E will function as the Right and Left triggers, R and F will function as the up and down on the Right stick.

That is much closer than what I expected existed, thanks. I'll keep an eye out for a sale. It seems to use Xbox controller drivers so it should be useful in Planetside 2, though whether I could get it to work in FL Studio is  another question. I wonder if that would mean I could translate that into playing synth guitar through an Xbox controller. That would be funny.

[methylatedspirit]

In order to test out video codecs, I want to extract a large (~1GB) amount of random bytes. random.org doesn't let you generate that much, not even through its white noise generator. There's no way in hell that a true random source would be able to generate that many bytes quickly enough, so I'll accept the fact that I'll have to use a pseudorandom generator.

What's the best way to do this? I own a Windows machine myself, but I also have access to a Linux box, so I can also use that if it ends up being more convenient.

[Starver]

Though it's not good for cryptographic security, you can probably script up something that runs an arbitrarily wide Linear Congruent Generator to push out bits/blocks of bits with long enough periodicity to get you the GB-worth of 'randomness' from anywhere within a potential TB/PB/more of its whole cycle (depending on seed). The mathematics of this is fairly well known (what to choose as a modulus, etc).

Or use a Linear-Feedback Shift Register like the XORShift method (easier to make the elemental binary work[1] without worrying about double-double-wide bitfields needing large relative-primes to be used in the mathematics). Just check the details on periodicity and make the state-width wide enough to add enough orders of magnitude above this to be happy. Then pump each byte (as you get it, whether as a trickle or a flow) into your ~GB reference file until it's large enough. First try it on a MB (if not KB) output, with the GB-capable algorithm, just to make sure you're not gonna wait five months. Shouldn't do (and if streaming to disc, the state-memory shouldn't strain your RAM capacity) but depends on how clunkily you implemented it.



Ummm... TL;DR;, look up LCG and LFSR pseudocode, and any notes on best use, and see if either tickle your fancy.


[1] I could probably hack up a very inefficient .BAT-based implementation, for at least the bitwise shuffling, but .sh, Java/Python/Perl and then actual compiled from a C-dialect/whatever would almost certainly be better with proper write-as-binary 'garbage-makers'. And if you're planning on sometbing like using the stream to flip raw pixel values, for codex-conversion/deconversion, you might be able to preformat it into the reference raw format of your choice and cut out the intermediate stage.

[wierd]

If you are willing to invest in some kit..

You can use such a PCIe to LPT card, along with a doubled up version of the circuit found here to get 8bits wide white noise on demand, at a pretty high sample rate.

You would attach the bit output lines from the 2 circuits to the D0 through D7 lines of a DB25 male connector, while connecting the common grounds to the grounds of the noise generator circuit, and then plug it into your LPT card.



(No, USB->LPT cables cannot handle bit-banging like this.)

From there, you could get a continuous, unbroken stream of random from that port whenever you read from it that is 8 bits wide. (one byte per sample)

The max sample rate of a high end parallel port is ~2mb/sec, so you are talking a good long while to get 1gb worth of random, but you can be assured that it is good quality analog random from quantum mechanical breakdown events occurring randomly in the zener diodes in your little dongle.  Alternatively, you could feed the quality random source to a pseudorandom generator as a randomly changing seed, to boost random value generation rates.

[methylatedspirit]

So, I have a lot of hard drives, and I want to repurpose the 1 TB ones (I currently have 3, but will soon have 4) as long-term storage. I don't have a PC I can stuff them all into, so I think I'll have to use one of those 4-bay drive docking stations to plug them into the tiny Intel NUC I have as a HTPC. I have two questions:

1) What RAID configuration do I use? I'd like it to be able to handle 1 disk failure (out of 4), and maximize space efficiency. Read and write speeds don't matter to me; it's long-term video storage.

2) What file system do I use? I daily drive a Windows machine, but I have intermediary drives that are compatible with Windows anyways, since it's a bit unsafe to carry 4 drives (some of which are 3.5") back and forth. The NUC itself is a Linux box. With that in mind, I want to explore other file systems that Linux is compatible with. Which ones are good for long-term storage, and are compatible with a stable software RAID config?

[wierd]

RAID 5, with striped parity. (Since you have 4 disks)

4 1TB drives will give you ~3TB of RAID storage, with 1TB of RAID parity data striped over the array.  The failure of one disk, is able to be overcome using the distributed parity on the other volumes. More than a single failure will down the array without chance of recovery.

I would use EXT4, with the proper RAID options set for stripe and stride.


Linux can do all this with software based RAID, using the [md] driver and toolchain.

[methylatedspirit]

My Logitech Keyboard K380 won't pair at all. It asks for a (randomized) PIN to be typed into the keyboard, but when I'm supposed to press "Enter" to confirm it, it does absolutely nothing. No response. I've tried this on a Windows 10 20H2 machine, Linux Mint 20.01 machine, and an Android 9 phone. What's unusual is that the numbers do actually register (as shown by the 'Mint machine), but the "Enter" key doesn't work. I've replaced the batteries already.