Every feature in RFF is a grudge
Let me get the embarrassing part out of the way: I have ADHD, and I am deeply impatient. Watching a spinner makes me want to throw a workstation through a window.
So here’s my villain origin story. I’ve run SCCM. I’ve run PDQ - which, somewhere along the way, also got slow (sorry). I’ve run Tanium, which costs like a small car and still makes you wait. Three “enterprise-grade” tools and the one thing they have in common is that they make me sit there. I cannot do it.
RFF is what happens when an impatient person decides “fine, I’ll build the fast one myself.”
If you’ve never heard of it: RFF is the Windows fleet tool - patch, deploy, remote in, lock down config - that I wished those three were. (How fast? I got our remote deploys from 183 seconds down to 26. But that’s another post.)
A little credibility before I start ranting
I didn’t come to this from a blog. I walked into a hospital that was running SMS 2007 - image DVDs and USB sticks, one WIM per hardware model. I taught myself MDT and PowerShell, stood up real imaging on my own, and built a desktop engineering team from scratch. By the time I moved on, that team managed around 30,000 Windows endpoints (and like 300 Macs, bless them). I was the first to roll out LAPS for our desktops - the other teams followed.
I tell you this so you know: every complaint below is something I’ve personally bled on. These aren’t features off a roadmap. They’re grudges.
Grudge #1: The laptop asleep at 2 a.m.
You schedule patching for the maintenance window. Great. Except half your laptops are closed and in a bag at 2 a.m., so they just… miss it. Forever. Nobody notices until the vuln scan lights up.
RFF keeps chasing them - an offline catch-up window so the box gets patched when it actually wakes up, not “never.” And there’s a check-if-idle option I’m weirdly proud of: tick the box on a job and it runs as the user, only rebooting a machine that isn’t being used. On a kiosk, that’s the difference between “graceful” and “you just rebooted a check-in screen with a nurse standing in front of it.” The user can even cancel it if they’re mid-task. Software that respects the human at the keyboard - wild concept.
Grudge #2: “Which GPO changed this?!”
I never got this fully solved in my hospital years, which tells you something. You know the drill - two machines behave differently, and you’re exporting registry hives into BeyondCompare, or diffing gpresult HTML files like a caveman, trying to find the one setting that’s off.
RFF treats config as something you can actually read and compare - capture a known-good baseline, see the drift, know what changed. Because I’m tired of playing spot-the-difference with regedit.
Grudge #3: Vendor silent installs (and the support reps who don’t know their own product)
I deal with software vendors constantly, and I say this with love: a lot of the support folks are useless on packaging. They don’t know the silent switches. They’ll swear up and down their app “doesn’t support silent install” - and then I’ll crack the MSI open in Orca and find the exact silent string myself (looking at you, Spok). Every. Single. Time.
This is the thing I’m most obsessed with fixing next: stop making every admin re-derive the same install strings in isolation. (A free public silent-install lookup is coming - because all of us figuring out the same strings in separate silos is a waste of everyone’s afternoon.)
Grudge #4: Intune, my beloved disappointment
We’re hybrid, so Entra already knows every device’s on-prem OU - it syncs right up. Can you build a dynamic device group on that OU? Nope. Entra flat-out doesn’t recognize the org-unit attribute for device rules. Your options are a short, fixed list of device properties plus 15 “extension attributes” you get to manually stuff the OU into yourself - to recreate something AD did natively two decades ago. It has the data. It just won’t use it.
And if you want a custom policy, here’s an OMA-URI string to hand-type like it’s 2009 - when you already invented GPO. Just… give me GPO for Intune. The deployment rings are clumsy, too. (Autopilot was fine; my buddy Rich at MS was doing cool stuff with FFU files - but with strict Epic naming conventions, Autopilot was never an easy fit for us.)
RFF does the obvious things the big tools won’t.
Grudge #5: Getting blocked mid-rollout over a license count
Few things are dumber than a tool halting your deployment because you crossed a seat limit during the rollout. RFF gives you grace instead of a wall - finish what you’re doing, sort the licensing out after.
Grudge #6: Dead machines you’re still paying for
A smaller one, but it adds up: dead machines pile up forever. SCCM technically has cleanup - buried in “site maintenance tasks” you have to know exist, monolithic and about as modern as a fax machine. ServiceNow’s CMDB just quietly fills with stale CIs. Either way your console fills with ghosts and your license count creeps. RFF cleans up after itself: a machine goes quiet, it frees its slot (reversible the second it checks back in), and actually removing it is a soft-delete so the history sticks around. Small thing. Still satisfying.
So why now?
Honestly? The dream’s been sitting in my head for years. The final straw isn’t one bad day - it’s that I finally have the leverage to actually build the thing I always wished existed, instead of just complaining about it at lunch.
RFF is for every Windows admin who’s ever rage-sighed at a spinner: the solo internal IT person, the small MSP, all of you. It’s fast because I can’t stand slow. It cleans up after itself because I’m tired of ghosts. And it does the obvious things, because somebody should.
It’s free for your first 100 endpoints. Point it at your fleet and see how “fast” is supposed to feel - getrff.com.