jthompson

Yeah, "Chromebook only" in that context would I think mean "this user setting won't apply on Windows, Android, iOS, etc."

When I navigate to Admin > Devices > Networks > [some OU] > General settings, the set of options there includes the "Blocked WiFi networks" option which I think you're trying to make use of. However, at the head of that page it reads Showing settings for users in [OU name], so I suspect that those options aren't actually applicable to devices, only to users. That doesn't quite tally with what you're seeing in chrome://policy on your devices, though.

[Android] Switch quickly between accounts in a Google app (e.g. Gmail) by swiping up or down on the account badge, without having to open the account switcher menu.

I've seen this with some old Netgear switches in the past. PING and SNMP response fails, so it appears as dead on the monitoring platform, although the device continues to perform as switch for connected devices okay. My guess was that the management part of the device would crash periodically, but the rest of it continued to work okay.

I've certainly found restoring entire VMs from Veeam nice and easy. I think I've only ever rolled back from Hyper-V checkpoints/snapshots when building some linux servers that I wasn't sure on the procedure for. Even then, that was years ago now.

It means that you don't have a Hyper-V checkpoint hanging around that would need a merge operation to get rid of. I haven't really used VM checkpoints for ages, so perhaps not best placed to properly answer your question. I guess I find the sea of green in the successful Veeam backup email notifications soothing and reassuring when faced with having to click on a scary button.

We use GLPI. It'll be able to do all of those things.

With incremental jobs, it does mean that you can run one on demand to quickly back up a server to that point in time within a couple of minutes or so, ahead of installing some big updates or clicking some other scary button. If you have plenty of resources, perhaps you could double up and have a daily incremental job, with a separate job scheduled to do a full later on. That's assuming Veeam is okay with multiple jobs for a VM.

Agreed that the email initially reads like it's a freebie that just needs enabling. £180pa for a CSV on demand. May be a no-brainer for plenty schools, but we already have a regime of encrypted offline exports on a weekly basis that are included in emergency grab bags.

I'd still suggest the VirtualBox approach, tbh, although there may be other ways to sandbox things (has anyone done this using Windows Sandbox?). It'll help to protect your hosts and wider network from abuse and accidents, and give you and your users peace of mind.

When you add a VM to a Veeam backup job, it will by default include all of the disks attached to the VM, even if they're not mounted in the guest OS at the time. If your VM was a file server that had e.g. 8 VHDX files attached to it, all of those disks would get included in the VM backup, and any disks added at a later time would automatically be included. Veeam just sees a VM with some disks attached, rather than working from withing the guest OS and seeing "OS C, Staff Share E", etc. If you wanted to only include specific disks, go to edit the Veeam job and in the Virtual Machines list for it, select the VM and click on Exclusions > Disks > VM > Edit.

Some useful pointers in this tread. We're considering moving from on-prem to SIMS Connected, and want to make sure that we have as many of our ducks in a row as possible. We're expecting sign-ins to involve a bit of waiting (our on-prem SIMS is not exactly speedy to launch as it is) but we'd like to get a sense if we can of whaat other schools' experiences have been like. How many of you have found it 'good enough', I suppose is what I'm asking. Would anyone be able to clarify a couple of technical bits for me? In terms of using commandreporter (part of the SIMS client installation) on an on-prem server, how is that instance of the SIMS client installed? Do ESS provide a simple installer which is then conifgured manually to connect to the hosted database (via VPN)? We absolutely want to retain the option to run commandreporter from Windows scheduled tasks on an on-prem server. It sounds from earlier posts as if that's possible, but I just wanted to be clear on how the client can be installed. Secondly, the SIMS ID documentation talks a lot about Google and SSO, but that's all about using SIMS ID as the identity provider to SSO into Google accounts. We would want it the other way around: having SIMS ID sign in automatically based on the current Google account. We don't want SIMS ID mananging Google account provisioning, and we'd prefer to not require staff to have SIMS ID credentials on top of their Google credentials. Does anyone know if SIMS ID can but configured in this way?

GLPI is definitely one to consider. The asset management is superb: very detailed and very powerful, especially if you deploy the agent. Either self-hosted or cloud option (https://glpi-project.org/prices/). The cloud version can authenticate against M365. Not used the helpdesk side of it, as I prefer Freshdesk for that (getting by okay with the free option).

Interesting. I'd be keen to see how well the racing sequences are shot. Hopefully it will steer well clear of any "I used to play Gran Turismo with your old man, kid. He was the best I ever saw" naffness.

It's like the Red Dwarf episode where a Polymorph consumes all of Rimmer's anger. "Call it extreme if you like, but I propose we hit it hard and hit it fast with a major - and I mean major - leaflet campaign." There's another choice line from that scene, but it's likely NSFW.

Secondary here, so pupils are placed in OUs by cohort and never moved. Are you able to set up your provisioning so as to populate pupil's 'Department' attribute in Google with their form name? If so, then some Apps Script might be able to routinely sort users into OUs depending on the value in 'Department'.

"Write a spoof news article about Uri Geller being trapped under a giant spoon." "Update that article to include some eyewitness statements and a statement from the emergency services."

Here's one dedicated to the wonderful mods here on EduGeek. "Write a story about EduGeek members, in the style of a superheroes comic."

If you can find any emails from [email protected] being delivered to 365 recipients, that may indicate that the recipient is a super-admin.

Yeah, I've asked it to write some Apps Script, and it used various bits of the Google APIs. That's obviously pretty well-sturcture information for it to ingest, but still, there's something incredible about seeing it happen. It's also good at apeing the grimdark, over-the-top style of Warhammer 40,000 (ask it to describe anything in the style of Warhammer 40,000). Getting it to write scenes from Friends is an interesting one. It knows the characters and some of their style, but there are never any punchlines.

Lat couple of rounds of phones have been Nokia, running Android One. Not sure what the deal is with Android One is these days, though, but hopefully the Nokia USP of a clean Android image with decent support lifetime is still there.

So, who's been playing with ChatGPT? I'm blown away by some of the output it produces. I feel compelled to share some of the things that have made me both laugh and gasp in awe at the same time. Here's a couple of repsonses to Tell me a story about Postman Pat in the style of hard boiled noire.

This. Shared mailboxes would ideally be accessed by end users via delegation to their indivudually named account, so that there's no extra sign-in required of them. Chrome/Edge profiles then for separation from any other accounts (i.e. personal accounts or accounts with external orgs).

Ooh, yeah. This thread on their foums looks promising: What are the best UK plugs with energy monitoring?

I keep meaning to look into this a lot more, a feel as though a suitable smart plug (something like this maybe - https://www.mylocalbytes.com/products/smart-plug-pm) that can provide wireless telemetry over MQTT might be the ticket. Perhaps a Raspberry Pi of some sort secreted within range to serve as a monitoring node.