TTS Tuff-Cam – “USB Device Not Recognized”

This is the TTS Tuff-Cam 2. It takes rubbish photos and is seriously overpriced, but many Nursery and Reception teachers have an aversion to giving 3-4 year olds a cheap Canon camera that they are normally perfectly capable of using, so insist on these “child-friendly” monstrosities instead.

Don’t buy one of these.

Against my better judgement, we bought one for our Nursery 6 months ago (and it will be the last one we ever buy). Occasionally, it goes on the blink and displays some or all of the following symptoms:

• Camera won’t turn on at all.
• Power LED appears to be stuck half-on.
• Windows displays “USB Device Not Recognized” when camera is plugged in.
• Other USB devices, such as the mouse, stop working completely a few seconds after the camera is plugged in, and start working again as soon as you unplug the camera (my personal favourite).

All of these have one simple cause: the camera is utter junk. Luckily, there is also a simple fix. Just stick a paperclip into this unmarked hole on the side for 5 seconds:

The camera will reset and probably start working again. At least, until the next full moon, or a butterfly flaps its wings near it, or something.

Screw it

Along with the near-certainty of personal injury and loss of fixings that comes with server racking, there is one absolutely immutable law of rack installation:

Any package of more than 8 cage nuts will include at least one that has no threading.

(and you won’t notice until you’re trying to put the final screw into the switch you’ve just racked up).

Not For Production Use

Recently spotted this while my home workstation was booting. It appears for a second, right before the POST screen, and is output by the MSI GeForce GTX 550 Ti I have installed:

What the hell, MSI?!

Now, I am become Death, the destroyer of… Printers?

Two weeks ago, the oldest laser printer in the school, a 12-year-old Color LaserJet 4500, finally packed up with loud grinding noises. Its final death throes included constant jams from tray 2, tray 1 feeding up to 50 sheets of paper at a time, and the drum drawer being completely jammed shut for 2 months.

Suffice to say, its time had come.

Since it was in a light-use location, I decided to swap in one of my Kyocera FS-C5100DN printers from elsewhere in the school, and replace that printer with the newer model FS-C5150DN. I’d bought one a few months earlier and been quite happy with it, so it seemed like a decent choice given my long-standing hatred of HP and more recent hatred of Xerox.

What a fool I was.

Indistinguishable

Today, due to a frankly ridiculous series of catastrophes, I have two different couriers picking up two different boxes for two different destinations.

The boxes look identical to one another, and contain almost identical cargo.

What odds can I get that they will take the wrong ones?

Why Portable Appliance Testing is important, you Idiots

Portable Appliance Testing (commonly known as PAT testing) is often spoken of with disdain in schools, cast aside as bureaucratic red-tape and disregarded along with countless other health & safety provisions. Teachers are particularly given to ignore it when they decide to bring in their own equipment from home, happily hooking it up to the mains without caring that it hasn’t been safety tested, because “they need it”, that all-powerful of reasons that many believe allows them to bypass any and all rules & laws (notably including copyright and data protection).

But I digress! Here’s just one good reason why PAT testing is actually important:

Last time these power supplies (all from old model RM Classboards) were checked, they passed. Now, time has rendered the plastic so brittle that all I had to do was grasp them and they literally fell apart in my hands. If they were plugged into a live socket, touching the now exposed innards would most likely give you a 230V electric shock.

“Ah, but AT,” I hear you protest, “I am not such a dumbass that I wouldn’t notice something in such poor condition. I would never allow such shoddiness is MY classroom!” Well, none of the class teachers in which these power supplies resided had noticed, despite them being in prominent positions in easy reach of pupils. The main reason is that most of them looked fine, until you either looked closely or handled them.

What’s more, experience has taught me that even an obvious fault cannot preclude a potentially fatal incident. Remember just now when I said that touching the insides would likely give you a 230V shock? I know that because two years ago, a teacher did just that with a similarly-damaged speaker PSU, despite admitting afterwards that she could clearly see that the live parts were exposed. Luckily she was not seriously injured, but the effects could have been more severe for a small child.

Hopefully this illustrates that PAT testing is not simply an exercise in box ticking. Failure to do it, and re-do it regularly, can put both staff and pupils at serious risk of harm. If it’s you that’s responsible for it, get it done. If it’s someone else, savagely beat them until they do it.

(By the way, if you’re looking for an alternative to the outrageously overpriced replacement RM Classboard power supplies, I’ve been using these, which at the time of writing, were less than 1/5 of the price. Which is nice, given that I need to replace more than half of mine this summer.)

Letters of Note

Something a little different today: as a follow of the often fascinating blog Letters of Note, I was thoroughly amused by this week’s post of a letter written in 1956 by Bill Hewlett, co-founder of Hewlett-Packard. I’ll give you a hint of its contents with the title of the post and let you read the rest there:

I have no personal knowledge of computers – Letters of Note

Please feel free to make the obvious jokes in the comments.

TODO: Hire QA Department

Sometimes I genuinely wonder whether some companies actually have QA departments. Earlier this week  I was once again suffering the inability of Adobe to actually post the right version of a download, and then today I was reminded of an old chestnut from AVG.

Here’s what happens when you forget to check your task list comments in Visual Studio:

Oh, AVG.

Terry is a lackadaisical idiot

Apologies straight away to any readers named Terry – unless you’re the Terry who filled in the Quality Assurance sticker on my server rack. Assuming he actually did the quality assurance, and didn’t just write on the sticker, that Terry needs to be shot.

My principal complaint with the aforementioned QA job is that about half of the locking pins that hold the doors and side panels on do not actually lock into place more than about 5mm, instead of the 15mm that the rest do, and they frequently only do that begrudgingly. This caused great consternation earlier today when I was trying to refit the side panel that is nearest the wall, not helped by the fact that it’s sodding heavy and there’s only about 20cm between the rack and the wall. I lost count of the number of times the thing dropped out because the locking pin not only hadn’t engaged, but had in fact come out of the mounting in the side panel entirely.

So, Terry, wherever you are, thanks for being a useless git. I hope you die in some kind of freak Cat5E cabling accident.

Why Acer servers are pants

Now that the new server room is complete, I was finally able to rack up the last server that I had left in its original home in the Bursary until the air conditioning was in place. In doing so, I was reminded of the various reasons why I hate this server, which is an Acer Altos G540.

First, a recap of an earlier rant: This server has badly configured default BIOS settings, and worse BIOS update procedures.

Second, the rack mount kit is a pain in the backside, like pretty much every rack mount I’ve ever used that wasn’t made by Dell. For starters, there were no instructions included with the kit. I eventually found them in the system user guide that came with the server itself. Next, I found that Acer decided to eschew the use of industry-standard cage nuts and require the use of a flimsy-looking shim that looks like a cage nut would if you ran over it with a steamroller. The manual calls them “distands”. I’m pretty sure this is a made-up word.

Handily they provide the exact number needed, though this is not so handy when one of them pings off into the void at the bottom of the server cabinet. You know, just like cage nuts. Except that I have more cage nuts, and no more distands, because no-one else in their right mind uses them. My server is now held up by only 7 of these stupid things; hopefully they are as useless as they look and the mount will be fine without number 8.

Thirdly, the server case itself is badly manufactured, something I encountered while trying to fit the rack rails. I was initially prevented from fitting the left-hand rail because a rivet that was supposed to be flush with the side of the case was in fact raised far enough that the rail would not fit. After a fit of swearing I ended up grinding it down with my Dremel, which while extremely satisfying, should not have been necessary.

Lastly, no-one stocks parts for the damned thing. The G540 is a late 2007 model (the school bought it in 2009 for several bad reasons before I was hired). Admittedly it is getting on a bit, but I can pick up spares for Dell servers from the same era (e.g. a PowerEdge 2950) from dozens of suppliers online. The only place I could buy the aforementioned rack for the G540 (p/n SO.RKG54.K01) was on eBay, and my attempts to find a spare hot-swap PSU (p/n SO.G5305.001) have proven utterly fruitless. This particular server has a 1-day fix maintenance contract, so the revelation that parts for it seem to be utterly non-existent is something that will make for an interesting discussion with the service partner next time the contract is up for renewal.

All in all, my assessment remains unchanged: don’t buy Acer servers. They are a waste of everyone’s time.