martixy

I hope linus realizes that this backlash is not over this ONE THING, but that it was the spark over many small things that have been building up. That's how this stuff works.

I hope linus realizes that this backlash is not over this ONE THING, but that it was the spark over many small things that have been building up. That's how this stuff works.

Interesting to see Steve guest-starring on LTT. 😄
 
Incoming GN merger? Buy more LMG stonks!
 

Interesting to see Steve guest-starring on LTT. 😄
 
Incoming GN merger? Buy more LMG stonks!
 

Please abstain from gaslighting here. "What you say happened did not happen, and even if it happened it was not a big deal"

It was down for 6+ hours. It was taken down by youtube the same way it was at the time of the hack. LMG did not disclosa anything about it. There is every reason for concern and it is not "ridiculous" or "silly".

Just theory-crafting here, but... Encrypting cookies with a hardware-based key?
Kind of like salting passwords in that regard.
The internet is decentralized, so there could always be local issues. Or things can break in unexpected ways. I have a channel that shows 404 for me, even though I can still watch videos from it.
But in this case it seems to be global.

This is getting silly at this point...
 
Nor do I see a thread about it. I'm never the first to notice, so... Could it be something local? The nearest google datacenter having a stroke? (Surely not threads getting removed.)

This is getting silly at this point...
 
Nor do I see a thread about it. I'm never the first to notice, so... Could it be something local? The nearest google datacenter having a stroke? (Surely not threads getting removed.)

Windows REQUIRES a minimal amount of swap space because its memory manager does not support overcommit.
 
It needs to be able to promise overeager or poorly written programs more capacity than they will ever require. This is the difference between system commit and physical memory usage: Commit=how much was promised, memory usage=how much of what was promised contains actual data.
 
Example: I on my machine currently have 27.3 GB of commit charge and 18.0 GB of physical usage.

Razor Basilisk Ultimate sound like a no brainer I guess, or Naga if you want even more bottom
 
I have never used Razor but my friend use their Naga for a long time and enjoy it very much - it definitely didn't last  long though, he solved that by buying a few extra Naga just in case they discontinue it or die off again. 
 
Imo, for a mouse with many bottoms, it's just inevitably that at least some of them will pose an issue sooner or later. Just go buy the design you like because in the grand scheme of things, nothing will last forever. 
 
I personally like Logitech for their subtle and almost professional look in their line up, when my G903 finally gave up its life I'll probably try G604 next - that or RoG Chakram, really like the concept although they still have less function when it comes to Logitech.
 
G-hub is all right for some hardware but absolutely terrible for others, and absolutely a horse shit garbage just a few months ago but I found that it has mature up a bit and not as unbearable to use any more.

Maybe I'm just unlucky but if the switches on both logitech mice I have fail in the same manner, and the internet reports the same widespread problem for the g602 switches, do you think that inspires confidence?

Just wanted to correct some misinformation that slipped into the latest techlinked.
 
Cellebrite is not a hacking tool. It's a tool for dumping and parsing the data of an already unlocked phone in a way that is convenient and easy for law enforcement to work with.
 
It requires physical access to the device, and requires the device to already be unlocked.
 
I CANNOT be used to hack encrypted devices and obtain your data.
 
Can you guys at least READ the articles before making news out of them? Maybe track the original source (linked below).
 
https://signal.org/blog/cellebrite-vulnerabilities/

I am a bit behind following the channel, and I just watched the Enmotus Fuzedrive video from a month ago.
 
This is the first time I've been able to detect this in an LTT video, but it featured a lot of misinformation as to how SSDs work, more specifically how NAND flash memory works. (Maybe because this an area in computer tech I know a little bit more about and I just miss issues on topics I know less about. Or maybe the writer fucked up.)
I was unable to locate any post addressing these inaccuracies, so I'll do so here.
 
The inaccuracies start at 2:25 in the video.
 
1. NAND flash operates on what are known as program/erase cycles. Technically the problem with endurance isn't the voltages required to "write", or program the cell, but the voltages required to erase the cell sufficiently quickly. This however is a minor nitpick.
2. "Multi-level NAND can be used in single-level mode, where only a single layer in each cell is actually used at a time." This is the major inaccuracy of the video - both what Linus says, and the graphic displayed. A cell is a single, atomic unit. It is not subdivided in additional layers. Multi-level cells refer to the amount of voltage levels that can be reliably written or erased from that cell. A triple-level cell needs to be able to reliably read and write 8 different voltage levels (3 bits, hence triple-level). As long as a cell can reliably store the charge and the controller can reliably distinguish between different charge states, a cell may be as many or as few "levels" as you wish.
3. "That means that it will sacrifice, over time, both the performance and endurance that it was supposed to gain from the SLC cache". Filling up a drive that can operate its cells in SLC mode for caching purposes, causing the NAND available for cache usage to diminish does NOT harm endurance. Endurance is based on the number of P/E cycles a cell has undergone, it does not depend on the mode it is operating. It does not matter if you write to a cell in QLC or SLC mode - you're used up one cycle either way. Furthermore, performance is not sacrificed over time (at least not anymore than with regular usage). Performance deteriorates over time with normal usage as a matter of course. A cell at a given wear level, but operating in SLC mode will inherently be faster than the same cell operating in QLC mode, because the controller can "push" the cell more. "[...] it can tolerate a higher voltage change until the erase will be so slow that the block needs to be retired"
 
 
Source: https://www.anandtech.com/show/5067/understanding-tlc-nand/2