psydii

You might be able to use openssl to create a root ca and codesigning certs? openssl - How to create a self-signed code signing certificate from a CSR? - Stack Overflow Deploy the root cert and code signing cert (without their private key) to the computers that need to trust the scripts using intune. keep the private keys well protected. Load the code signing cert into the private cert store of the user account who has the authority to sign them. Make sure you have a way to revoke the certs though.

and yet... **Stares at Bromcom's reliability problems.** Honestly, using SIMS NG actually primes a wide range of stakeholders to ask the right questions when you approach the end of the contract and need to re-tender.

There are only two things that are inevitable, and both will be involved when the time comes.

Nope. Didn't say buy a rubbish product. I meant: if you've got SIMS NG, do the world a favour and help them make it viable against Arbor and Bromcom. We will be in a very bad place if in five years the *only* choice is Bromcom or Arbor.

FWIW I do think that everyone with SIMS should be making use of SIMS NG at least to some extent. It is still fully dependant on SIM7, so it can be viable even without being anywhere near feature complete. We need 3 viable options in the MIS space, and SIMS NG needs to have users in order to develop into a robust product.

Have they not heard of wrist watches?

Yeah exactly as @msi_school, you might want to check RCC actually work in your location. We had a tonne of trouble with ours and ended up going back to manual. But these days most people just rely on the clock on the computers.

We've had a few epson surecolor's. The smaller photo printers all seemed to die for one reason or another. But the big floor standing "technical printer" (which is a actually used mostly for posters and photos) has lasted five years so far. Photography dept use it by request, usually only for the shows/exibitions/moderation day as they've moved the workflow including the portfolios to a digital platform a few years ago. We also have our big repro colour machines spec'd to produce good quality photo print on standard a4/a3 so we can produce some marketing and induction materials in house at close to the standard you'd get by outsourcing the printing. These see a fair bit of use from the photography department - higher quality than a standard 'colour capable' laser or photocopier, lower than the bit wide form, but used to be the primary devices they used. The team were very happy with them.

It's the price hike they warned they'd hit you with if you didn't agree to a enter into a new 3 year contract by December 2024 (even if your renewal is in April).

Docklands to Hampton Court is a *trek* . Might have been quicker to go by boat down the river, but then you'd have gone under rather than over Tower Bridge. Sorry to hear how badly you mashed your hand, but I can confirm that were able to type legibly.

Pretty sure they are only just limbering up. Once Bromcom isn't owned by its founder (ultimately inevitable in my opinion), they will raise their prices and this with untether Arbor and SIMS, enabling them to do what they need to for the ROI expected by the PE funds that own them. Honestly I wouldn't be surprised if by 2035 schools end up paying 10x what they are now.

They are in the MIS and SCR, so they get accounts provisioned, so they use Teams. They also have to use the wifi because 4/5G coverage is patchy at best. We also have a no-phone policy. But its implemented much like the pirate's code.

...while the lifetime warranty on the 2920's we have has done some seriously heavy lifting.

We're old fashioned* and just join them to the domain and let windows / entra connect to do the hybrid bit. A if very basic WCD profile does the initial domain join during OOBE. We have SCCM handling the enrollment to intune for now. But if you don't have that, you can do it with Group Policy Enroll a Windows device automatically using Group Policy | Microsoft Learn We don't have these devices configured for Autopilot. My understanding is that if you have autopilot devices, then you need special profiles to AD join them as part of the OOBE. *I used to be all about the clean built vanilla image, but these days with Lenovo Thinkpads and the Advantatge software, "as they come from the factory" is actually better and less work for us. Experiences on other hardware may vary.

I've tried to keep things as simple as possible over the last decade or so. Previously we had non-stacking switches at the edge, patched together so no single failure would take an entire comms room out. The problem was any single switch failure would cause a significant spanning tree event. By putting a stack at the top of each comms room rack (top switch to core, bottom of the stack to the next comms room around), and then hanging the the legacy switches off that stack with LACP Trunks network reliability improved massively. If we'd put in 24-core (rather than 12-core) OM3 between the comms rooms when we refurbed the building we could have had every switch with its own dedicated back-haul to the core, the benefits of stacking (in a school environment) would have been somewhat redundant. I have also seen entire stacks crash (so much for improved resiliency!) but these problems have always gone away over time with firmware updates. Our core at top of rack stacks received their last updates *14 years* after initial installation. My concern for lower cost solutions is whether the product will be around/supported long enough to work through the bugs, and how difficult the sell to SLT/Governors will be that we need to replace everything again after only a few years, and that the new kit has to go through the pain of being new. Honestly, if HP/Aruba said "give us 10K per year and well support your switches for ever until you stop paying", I'd be the first to sign up.

If the Gen 2 docks overlapped with the faulty laptop usb-c/thunderbolt intel hardware/firmware from 2018/19/00, then them dying suddenly was likely a function of the contagion from that disaster. We had many gen 2 docks fail 2021-2023. More recently we've been deploying previously unused Gen2's and ThinkPad Universal USB-C Dock, and they've thus far lasted longer than the ones that had been exposed to the older laptops, but not yet at scale. That said, what docks did your desktop team settle on after the Gen2's? (We've found the Cable Matters USB 4 cables to be reliable and have replace only a couple of the 20 or so we've deployed in the last year and a half, and those had clearly been mistreated)

@MrEprise who ever mandated the change should be made aware of the hard data from the helpdesk showing there is a problem. Explain what you’ve done thus far to fix things. Then (crucially) in the spirit of being seen to be supportive of the senior person who “sponsored” this change, you offer three solutions: plan a: ask them to authorise spend to try a room or two with the cabling @Mr.Ben recommends, and if the doesn’t improve things, try also the specific dock he uses. Make sure that laptops used in these rooms are fully up to date with drivers firmware, os updates etc. plan b: arrange things so you have hdmi direct to the laptop. (You may still need to upgrade the AOC @Mr.Ben recommend) plan 😄 revert to desktops Also, float the idea of calling in central IT to review/consult since they exist, thought it was a good idea and sometimes problems like this are bigger than one person can resolve. To be clear: I moved to usbc docks. It was a disaster of similar scale. We ended up going direct hdmi (with aoc) . We’ve since moved 10%back to docks (newer laptops, newer docks, new cabling as per @Mr.Ben) and docks are now fine. Laptops are never as reliable as desktops, but the trick is to make them reliable enough.

Our 5 minute average on the wireless today has hit over 900Mbps, which I think is a record, but not the full story. The edge firewall actually matches pretty closely to our wifi dashboard: Not sure which is more significant in supressing the numbers, either TCP congestion management and wifi collision domains (which to over simplify TCP will keep backing off every time packets get dropped, and packets are dropped when a link is saturated or collisions happen (which back when TCP was invented, were basically the same thing). So it may be that the busy wifi network (wifi 5 here remember) will be suppressing traffic rates, when we'd intuitively expected to see many more peaks nearer 900Mbps) Or the fact that this is averaging out over 5 minutes there might be a 30 second peak of 900+ and then is drops to the background 250Mbps again, which averages over the five minutes at well lower than the 900Mbps users actually experienced. There are a few 'flat top' peaks in there, which suggest we are hitting internal through put limits on either our internal-vm-based-not-optimised-at-all firewall or wifi - it shows up in the uploads 200Mbps is about the fastest we see traffic on a busy ap so and we had a class uploading photos at about 11:30 today which coincides with a flat-topped peak in the outgoing graph.... The Download peak around 9am also has a bit of a flat top, and again was likely a couple of AP's serving the same photography class - higher because wifi 5 is much beter with dowload traffic to multiple clients than it is with uploads.

Talk to your CSP, they can usually navigate the vagaries of loading in the correct licence.

...and with 30ish laptops on the AP do you ever see more than 1Gbps wire-side? (I know you will with only a handful of clients, but once you load them up can they sustain the throughput?)

Yeah, I'd say now is not the time to be looking to rebuild that, maybe go for a gpu with 12/16Gb so you can get AAA* graphics at 1080p on older games or AA* graphics with Ray Tracing on more recent games. Take the money and put it in a savings account (if you are fiscally conservative) or a tracker account (if you like rolling the dice), top it up a little each month and then revaluate in another couple of years. Don't worry about the CPU being a bottleneck - there always has to be one somewhere. Maybe if it bothers you enough, look for a second-hand socket compatible, i7 or i9? (but if you go this path, ARC is off the table I think because of a lack of rebar support on your mb/cpu... hmm do modern nvidia's need rebar to get their performance too? maybe you do need a more modern cpu/motherboard...)

The real problems* for Bromcom users will start when the founder/owner passes. This problem will be contagious and all schools no matter their MIS will look back on these times as a golden period. The problems will not have a technical foundation but a financial one, one already laid deep in SIMS and Arbor. *not in anyway to dismissed the very real problems users are having right now, but right now they can move for a reasonable fee.

Thanks for posting that. I saw it cross my feed last(?) week. but didn't get to it before it dropped off into the oblivion of things I'd meant to do. It's worth looking at where MS is with Windows and Azure, and look at how they recovered from the Cairo, and Longhorn fiascos. On both those occasions the basic gist was Dave Cutler (an engineer "without portfolio", who reported direct to the CEO) had a skunk works team who maintained a modern build of his OS (for he is/was to the NT kernel (and for a while the original RedDog Hypervisor of Azure and XBox) as Linus Torvald is to Linux), that actually worked despite whatever the mainline PMs and other engineers were doing. It was that branch that was used as the baseline when Cairo and, a few years later, Longhorn got cancelled. Mark R took over for windows 7, but then moved to Azure (possibly with a detour to write stuxnet *allegedly). They were both the sort of people whose direct reports (if not themselves directly) would have shaped that engineers outlook on system stability, matching the low level engineering efforts to hardware, and the needs of the customer further up the stack. Cutler is 85+ and Mark R star waned (publicly anyway) after a big multiday Azure outage about a decade ago. It's time to empower these sorts of engineers again, those who understand what they have, what customers need and what needs to be done to deliver. (I wrote this before reading part two. lol. yeah.)

I feel that there are parallels here with Bromcom. They both promise too much to too many, trip themselves up and frustrate everybody. SB and Bromcom both appeal to *me* with the flexibly and control they offer, yet here I sit on LGfL and SIMS because I need to be able to sleep at night.

I concur regarding Lindy and Cable Matters cables linked. We use Bluestream for our HDMI cables (which are so good it feels like the HDMI protocol/speed negotiation happens just before you decided to plug it in), but they are twice that price.