The reason for this I heard by Microsoft was to stop having to explain to non technical people that Azure AD doesn't have feature parity with a real Active Directory.
As a Portuguese native Microsoft FTE, I found the name change jarring. “Entra” means “Enter”, with a subtext of permissiveness that seems overly positive for an authentication mechanism. But then again, Continental Portuguese is seldom considered as a semantic idiom, and Brazilian Portuguese (or even Spanish) tend to have a lot more influence in branding everywhere on the planet.
There is an experimental flag[0] in the .wslconfig which let's you set autoMemoryReclaim to gradual or dropcache. On my laptop I had good success with dropcache and on my countryside desktop with gradual.
As someone using both equally often, I am kind of certain Android passes through the raw sensor readings while iOS has some form of debounce delay and guesstimation implemented. Both are equally often off in their readings. But that's just based on usage experience, not a hard fact.
I don't recall the exact context anymore but during the Microsoft Build this year it was said by an employee at Github that the service came with the acquisition which is the reason it was and stays C#.
Zerotier maxes out based on the model since it's single core. On a ac3 around 20Mbps is the cap. With wireguard I max out at 300Mbps. I can't test openvpn since mikrotik doesn't support tls auth. IPsec was always my given linespeed.
Beware that you need fasttrack exclusion rules (via connection mark e.g.) for each vpn interface except zerotier.
Well the problem arises if you use more than a few shortcuts.
If you start having keybinds like ALT + CMD + (F1..F6) you _really_ dont want to have to press an additional FN Keys. I found 4 key shortcuts not viable in the long run if you have to press them multiple times a minute.
With that said:
> So for me personally it's a number of advantages without a single drawback
The objective drawback is the expandability of your shortcuts because you are artificially limited in your options of key combinations. For me that would be a no go. Though I am close to 500 shortcuts in my 3 main applications alone, excluding i3 or KDE globals. I only use a mouse to browse the web.
If you have to regularly enter 4 key shortcuts you can just
define a few macros in the firmware and be done with it.
I haven't been in a situation where I absolutely couldn't input
some combination. Some shortcuts are awkward, sure, but I rarely
need those.
If I had to regularly use 4 key shortcuts I'd go insane no
matter what keyboard.
Edit: Then again I don't (yet) have 500 shortcuts, but if
software doesn't support rebinding those that's not a problem
with the keyboard imho.
Those keybinds are mainly custom macros and scripts.
Some project specific some more general. The project specific keybinds range from do test deployment and run tests to grab the latest test results format a text variant as well as pdf, archive and send them out to everyone who needs them.
The general ones range from tell my AC/heating that I'm heading home from office to open my local weather forcast on Screen #2.
IMO there's no good reason to omit possibilities for aesthetics. Though as often as I thought about it I am not going to use a stream deck since I know I couldn't live without one after a few weeks and don't want to be bothered to carry it with me while traveling.
> IMO there's no good reason to omit possibilities for aesthetics.
My point is that it's not for aesthetics but comfort. For
aesthetics I wouldn't use vertical stagger or a split layout.
With a 40% keyboard you can enter everything without moving
hands or wrists at all and more important keys are moved right
below the thumbs so I can use Ctrl and Shift without twisting my
hand.
All said and done I still think that the standard keyboard
layout is actually not a bad choice either, and is sometimes underrated. I just think that
some more experimental boards manage to be just as productive
while removing some ergonomical weaknesses.
That is not an argument for more non-modifier keys; it is an argument for more modifiers. Why? Each added non-modifier key adds X more modifier+non-modifier keybindings (e.g. C-n) and X-1 more modifier+modifier+non-modifier keybindings (e.g. M-C-n), where X is the number of modifiers. In practical terms, a keyboard with control, alt, shift and super only gets four two-key and three three-key bindings with each additional key. That is not great.
Adding a modifier, OTOH, add Y modifier+non-modifier and Y modifier+modifier+non-modifier bindings, where Y is the number of non-modifiers. So adding a hyper key means adding , say 104 additional keys (on a keyboard with 52 non-modifier keys). Adding a compose key means adding … another 104 additional keys.
My current keyboard has control, alt, super, hyper, compose and shift. A lot of folks add raise & lower, for even more modifiers.
For many, additional modifier keys means keyboard firmware changes with macro programming. My keyboard does not have media keys, but I have programmed the keyboard that Fn+Q sends the Volume Up key, Fn+A sends Volume down, Fn+W sends Play, etc. There is no Fn key in the HID specification (at least not that I'm aware of), the Fn key exists only on the keyboard's firmware.
Many keyboards these days are getting quite fancy. There's even an open source keyboard firmware standard.
https://qmk.fm/
I use X11 and a custom keyboard firmware: both of my hyper keys generate USB left super and both of my super keys generate USB right super (or vice versa), and X11 turns left super into hyper and right super into super.
It's not a scam. Those scenes live by reputation alone. If you need something as a once in a lifetime service you are well advised to use a trusted third party within the scene. Those are either trade mods or well known veterans.
If you become familiar and known in the scene the risk of being scammed is very low and if it happens it's more like a "one last money grab and I am done thing" where the person offering the service will disappear. But since this works once per online persona this really doesn't happen that often.
how does one even begin to become known in a scene if you don't even know where (or if) it exists?
Not that I'm planning to purchase any of those services of course, I'm just curious because it sounds like there's no possible starting point, unless by pure chance one of your personal friends happens to be already involved in the area and lets you know.
Dark web markets have escrow systems as well as review systems which show how much each customer paid for the service. One can see if a vendor is well reviewed and it's unlikely the reviews are astroturfed if they are on large transactions because the fees on these markets are relatively high.
you start working through an established escrow service, and gain positive feedback on darknet platforms.
You can also do what above-ground companies do and promote yourself via ads, branding and so forth. At the very least that signals you've invested serious money into your image and are therefore unlikely to disappear overnight.
You pointed out the possible starting point yourself: one of your acquaintances is already involved with the scene. That seems quite sufficient for the scene to keep existing.
Online reputation is very easy to build. These people are scammers they're hardly going to worry about giving themselves good reviews. It's done all over amazon, trip advisor etc yet somehow the darkweb has a full proof online reputation system - don't make me laugh.
Dark Web reviews are linked to purchases. you can't just had a load of fake reviews without completing a transaction which means you have to pay a fee to the darknet site. So it is possible but it can be expensive, the easy option is to run a legitimate service and then at some point you could just start taking people's money and not providing the service. Of course this has a limited shelf life, because new reviews are going to be a positive
Tried that with a fully specced out machine. It doesn't matter if VNC, X11 XDMCP, Teamviewer, Anydesk or some of the more obscure things I tried out. The latency kills all the fun.
I've found that while it changes the constraints on what is fun, some fun is still possible, and the ability to preserve state across devices/locations sometimes offsets the 'loss of fun'
e.g. coding / light 'ops' stuff that is just 'graphical text' is fairly bearable to me in the ~100ms range with good compression/low bandwidth settings, since often you can blindly type some code/commands and take a breath while the UI updates. Video/Graphics/Etc or heavy GUI interactivity starts to get a bit painful.
one clear benefit to cloud is faster in-cloud bandwidth/latency which could make up for the less responsive UI depending on your use case
Obviously you can get work done with it, I also can imagine that getting accustomed to a type and think about it approach while taking a breath might be worth the ability to preserve states. For me preserving states across devices/locations has been solved 10 years ago, though nowadays in extreme situations I just don't do any work if it's not my device. I won't enter any credentials on a device that's not mine or hasn't been in my physical control since being setup. I'd rather fly home. Nothing we do is that important.
For me a low latency is the single most important thing when using a computer. Having 1000hz input device rate on the mice, proper NVMe SSD and 144hz/240hz Display rate makes all the difference of a decent computer experience.
Several years ago I had pretty good luck with the NXMachine server/client for a training system. I don't know if they still offer a free version or not :(
Yeah, NoMachine NX, used the free version ~2010, it was lightyears ahead of any other remoting solution for Linux. Presumably, the commercial product still is but there is no free version anymore.
There was a freenx project but it seems not much alive now. A different open-source project called x2go continues. The implementation seems less complete and stable but nevertheless, still a decent option for accessing a Linux desktop remotely over a lower-bandwidth (e.g. 3G mobile) connection.
