Youre talking about Linux containers on Windows, I think commenter above was referring to windows containers on Windows, which is its own special hell for lucky folks like me.

Otherwise I totally agree. Ive done both setups without docker desktop.

Windows container runtime is free as well, simply install the docker runtime from chocolatey or winget along with the Windows Containers and Hyper-V windows features. This is what we do on some build machines for CI.

Theres no reason to use desktop other than “ease of use”

Oh I’ll have to check that out I thought I read something about that method being patched.

Tho I do like just booting a new install and its already activated automatically :P

if you’re in the know, check out vlmcsd on github and “test” windows enterprise with KMS. It can run on everything from a pi, to docker, to openwrt. If you’re really gangster, you can set up SRV records and get auto activation on your lan

Holy shit, that’s actually hilarious, I imagine someone would have noticed when their paste/auto type password managers didn’t work

For those confused, this sounds like instead of making a real website, they spin up a vm, embed a remote desktop tool into their website and have you login through chrome running on their VM, this is sooooo sketch it, its unreal anyone would use this in a public product.

Imagine if to sign into facebook from an app, you had to go to someone else’s computer, login and save your credentials on their PC, would that be a good idea?

I think you could boil it down to something like Set-ADUser bob -otherattributes {uidNumber=1005, gidNumber=1005}

sorry I don’t have any real documentation but I have a snippet of powershell that explains it pretty well here this comes from a user creation script I wrote back when they removed the unix UI.

I was using Get-AdUser and discovered that the properties still existed but you have to manually shove those in, when an sssd “domain bound” linux machine has a user with these props login, they get the defined UID and GID and homefolder etc.

$otherAttributes = @{}
Write-Host -ForegroundColor Yellow "Adding Linux Attributes"

# get the next numeric uid number from AD
$uidNumber=((get-aduser -Filter * -Properties * | where-object {$_.uidNumber} | select uidNumber | sort uidNumber | select -Last 1).uidNumber)+1

$otherAttributes.Add("unixHomeDirectory","/homefolder/path/$($samAccountName)")
$otherAttributes.Add("uid","$($samAccountName)")
$otherAttributes.Add("gidNumber","$($gidNumber)")
$otherAttributes.Add("uidNumber","$($uidNumber)")
$otherAttributes.Add("loginShell","$($loginShell)")

$UserArgs = @{
    Credential = $creds
    Enabled = $true
    ChangePasswordAtLogon = $true
    Path = $usersOU
    HomeDirectory = "$homeDirPath\$samAccountName"
    HomeDrive = $homeDriveLetter
    GivenName = $firstName
    Surname = $lastName
    DisplayName = $displayName
    SamAccountName = $samAccountName
    Name = $displayName
    AccountPassword = $securePW
    UserPrincipalName = "$($aliasName)@DOMAIN.COM"
    OtherAttributes = $otherAttributes
}

$newUser = New-ADUser @UserArgs

basically the “OtherAttributes” on the ADUser object is a hashtable that holds all the special additional LDAP attributes, so in this example we use $otherAttributes to add all the fields we need, you can do the same with “Set-Aduser” if you just wanna edit an existing user and add these props

the @thing on New-ADuser is called a splat, very useful if you’re not familiar, it turns a hashtable into arguments

lemme know if you have any questions

Microsoft pulled those from the UI, but if you’re adventurous you can just shove those attributes in to user with power shell and it works the same.

Then just use sssd instead of NIS, surprised me at work when this worked.

What a great series that is, I should get the kit

Im a security professional who works to harden medical devices. I use the flipper zero to easily test many different protocols that would be a pain in the ass to do “manually”.

The flipper makes it easy for me to verify IR, sub GHz, USB, SPI, and many other protocols while being able to walk around the devices I test.

Without the flipper I could totally do these checks with homebrew tools, a pi and an rtlsdr (unless thats gonna be illegal too?) But it would take me writing new tools and procedures rather than the ease of the flipper.

Anybody in the know can tell you that the hardware isn’t anything special, and like many others have said, its like making a swiss army knife illegal cause the toothpick can be used to pick a lock.

This isn’t gonna stop anybody, if pentest tools are showing flaws in your product, maybe we should send flippers to the car manufacturers and tell them to fix their shit. You shouldn’t be allowed to sell a car that can be wirelessly hacked like this, just like how the FDA doesn’t let you sell medical devices that can be hacked like that.

You don’t just put the cat back in the bag…

If you’re mixing a dedicated GPU and onboard graphics you need to set the dedicated GPU as primary somewhere, otherwise all screens get rendered on the onboard and “reverse PRIME’d” to the dedi GPU outputs.

I’ll see if I can find the snippet that fixed this for me.

This also adds the benefit that any other devices that wanna VPN can just use the gateway

I’ve read some of the comments and it sounds like you’ve already tried installing proton VPN and tailscale on the same machine, but depending on your setup maybe you could make a “VPN gateway”

Like take your pi, install protonvpn, then enable IP forwarding and use a little nat IP tables script to nat your lan to your proton VPN interface like a home router would with the wan and lan ports.

Then on your tailscale gateway set the default route to be that box instead of your normal router. Then just use the tailscale node as the exit node on your client and check your IP.

In theory this would be similar to a qubes type setup which is what I tend to use for this kind of work.