psydii

I vaguely recall having a problem like this sometime 2001-2004. I think I ended up just pushing a .reg onto the computer / user that set it the way we wanted. What happens when a printer is added to Windows, how the server-side settings of the printer impact client-side settings, and whether Windows (and the print components from the driver) can see the printer's management interface (and what that management interface actually is (usually SNMP but not always?)).... is woefully documented. I think there used to be Technet or Technet Magazine articles, but they are long gone. Then there was the v3 v4 driver models and then Print-Nightmare thing messing things up further. To be honest I think there is a chance the only people who might be able to document how windows printing works in various scenarios are the team over at Papercut, but they probably have signed NDAs beacuse the whole thing has more gaps in it that swiss cheese.

It's a breach, but not one that requires reporting to the ICO. Personal data breaches: a guide | ICO

Is there an app that has custom print settings causing it? SIMS, Acrobat, Word IE (if its still lurking deep down somewhere), Illustrator, Chrome/Edge would be places that might be worth looking at. The default profile on the impacted computer, and the network default profile seem like other places to go splunking if most existing users are not affected, and it impacts a user even after a

If we’re going to meme it, SIMS is more closely tracking the Half-life 3, timeline, they’d even got their own equivalent money printing machine to Steam (ParentPay). However the quest for nuclear fusion power plants is just modern day alchemy.

....and the understatement of the year award goes to...

Yup us too. We use only a tiny subset of its features, which seem to be working well enough, but was intrigued by the version.

Just out of interest, in the help-about menu - what version does it say you are running?

It doesn't require on site server, but does require a sims.net server installation somewhere. Also you still need the .net client for at least the data team, timetable team, the cover team the SEN team, and anyone who uses any of the sms/email messaging services. So its not directly comparable to Arbor or Bromcom yet. It is moving quite quickly now to bring the core functionality to a modern web interface. I'd say likely 3 years and it will be a full MIS in its own right. Now, will they have any users left at that point? It's hard to say. I was surprised that the numbers hadn't collapsed to single digits at the last review of census data (where an FOI can get us the data on which MIS schools are using), so maybe they will make it to full cloud this time (its their third attempt in the last 15 years).

This is more an informational post to future people searching for solutions to USB C related issues, but it may be worth considering - at some point continuing to replace dongles might be more expensive that an actual dock. USB C is a fickle beast. Cables are fragile, and if they are not replaceable but the device they are connected to are in effect "consumable". But then if they are replaceable, that's an additional physical connection that needs to be "good" and the ongoing danger that a replaced cable is sub-standard. How Long Can a USB-C Cable Be? What is USB-C Power Delivery? What is USB4? USB-C Cables and USB-C Adapters from Cable Matters Also dongles are not designed with the expectation that they will be used by 6 different people every single day, where as docks (particularly post-covid where hot-desking is much more common) should be and more engineering effort has probably been spent in protective mechanisms for the electronics. Anker is a good brand, and with a 5gb/s 4k/30Hz cap the are probably using USB 3.2 Gen 1 or 1x2 rated cabling which allows them build a product that is not quite so electrically sensitive to wear and tear. All our staff laptops support DP-Alt mode, so where we have dock/dongles we install a full USB-C 3.2 2x2 Dock with PD and DP-Alt mode we now use only USB 4 USB-IF certified cables.

Same. Its a split. Though we now hold the spares for card readers and controllers and cameras and tills, very much not my budget though. The day to day admin of the services is handled by the office manager's team. I audit them wrt CCTV access, the business manager audits them on access control cards/cashless catering provisioning. End of year processes have just moved into the IT Office, but that is due to a specific staffing structure rather than it being where it should sit in a more general case.

No complaints from colleagues, but I've been hitting the cloudflare "are you a human" screen most days when trying to access edugeek.

If it took four days to do all that, I wonder why they didn't spent any four days *before* the start of term doing that? Start of term work-loads are 100% predictable. This was avoidable. (Mostly. Instances failing under load due to flawed implementation is testable/avoidable, failures in the underlying cloud providers infra-management (recycling azure components), less so - but testing their own code properly would have enabled them to avoid entering that failure state/path in the first place.

Salamander told us last year that SDS has changed and now provisions the members explicitly as "members" not "subscribers" overriding whatever it set at the team level. Their only fix was for us to give them a GA account. So instead we just run a script to made all members subscribers. $classteams = Get-UnifiedGroup -Anr "Section_2024" foreach ($class in $classteams) { $class.PrimarySmtpAddress add-unifiedgrouplinks -identity $class -links (Get-UnifiedGroupLinks $class.PrimarySmtpAddress -LinkType member) -linktype subscriber set-unifiedgroup -identity $class.PrimarySmtpAddress -AutoSubscribeNewMembers:$True } You need to rerun it when class memberships change (or at least you did last year). There's probably a way more efficient way of doing that.

To me, its not so much the architecture of the systems, its that they don't appear to have tested against predictable workloads and then act all surprised when the service buckles. At the end of the day, whether its one sql server or hundreds of containerised microservices, ultimately its all running on shared cpu/memory/disk. If some common transaction, or underling piece of hardware that every school's instance uses becomes non-preformant, everybody feels it.

Consider that a win. Last time ofsted were in I fixed the coffee urn that we'd put in their room. It didn't even have a plug!

I miss Becta FITS being a thing that those going through the NPQH used to learn about.

Is there any chance there are members of a group that has and exclusion rule applied to it? Are they members of more than one group and both groups have separate policies that apply? You (I think) need to exclude the group-you-want-your-policy-to-apply-to from all the other policies.

I saw this exact symptom with a different vendor's dock two days ago, and replacing the laptop-to-dock usb c cable with a "USB-IF Certified USB4" 1.2m cable, *and* making sure the dock-to-board usb connection was via one of the 20Gb/s "USB A" 3.2 ports on the dock, so that electrically it was less likely to actually be daisy-chained off other hubs inside the dock. Strange thing was prior to having this problem, the cabling/dock/laptops had not complained about "too many hubs".

How are you building your user profiles? We don't use a network based one, and each is created on the local machine at the user's first logon. We use GPP to ensure that [empty] "OneDrive - OrgName" and Documents etc sub-folders are all pre-staged the Default user profile, so when OneDrive fires up for the first time in a new user profile the folders are in the expected places. We also still use GP to redirect the user's Documents etc, to these folders inside the user profile "OneDrive - OrgName" etc. folders. All the OD client needs to do is start syncing. On "sign me in to all apps" I don't think you should not be seeing that in an AD / AAD Hybrid Join environment. Something isn't quite right there.

There’s too many variables these days. Laptop chipset, firmware, and drivers, the actually robustness of the usbc port in any given model, the usb c cable quality (even if you get one that meets the required standards, it might not be robust enough for 4-10 hot-dealing events each day), the dock model, the dock’s own firmware(s) (each group of ports seem to have their own firmware updates, though some dock manufacturers obscure that fact), the hdmi cable quality, the board port quality, the board’s own firmware, both at the Android level and that actually running at the hdmi/power level. Sometimes you get lucky and you’ve deployed a working solution, sometime you find next year’s laptops won’t work right, sometimes the cables work for a while, but then become unreliable, sometimes it’s the docks. Sometimes it all works fine and then a firmware update hoses you. Generally it does seem to be getting better, if you have a reliable cable supplier, and your board and laptops are in warranty, we’ve found glitches resolved with new firmware and the vendors already have it when we call. Having docks in the mix adds a whole additional level of complexity and it’s been not unheard of for dock vendors to simply EOL the dock rather than fix it, though again this has improved somewhat in the last few years. We recently found some docks started working with a whole class of laptops that we previously never reliable with docks after we replaced some of our 2023 cables (which at the time we were over the moon with as they worked perfectly 100% of the time with 60% of the fleet) with our 2025 cables. Why did we replace them? Well because they had become damaged. How did we discover they fixed the problem with the 40%…? Well, a young teacher ignored our instructions and tried the usb-c only lifestyle, found it worked and told their colleagues and eventually we caught a teacher doing it the wild. We are watching with trepidation, maybe this time, maybe this time…

Does anyone have a real handle on the scope and scale of this incident? The back-end provider seems to have been used by several services, across several sectors. It feels to me like it could be somewhere between everyone with a dbs check, or only a few thousand people. Anyone found they are on the list? (though maybe we don't want to say, just in case cross referencing the users from here against the list provides further advantage to the bad-guys)

Well, I was kind of expecting someone else to jump in after that rather unhelpful narrowing of the problem. @StephenPink is this still plaguing you?

I'm not saying they are losing hearts and minds, but out classcharts "champion" has taken a look at sims next gen, and openly speculated that it might be better value.

Well its not an intune policy that is the problem them. Hope this helps. Best of luck.

No intune involved?