So I’ve had three main projects over the past few months: build a new test rig, get a Virtual Reality rig set up and test the latest bunch of MU-MIMO routers. Has any of it worked first time? Of course not.
This job is great in that you get to play with the cutting edge tech and gadgets but it’s bad for the blood pressure. In an incredibly time poor industry, there’s not much time for testing as it is, and so when something doesn’t work right away it snowballs into massive delays across the board.
We have an epic test rig all ready to go… in theory. However, when we finally put it all together, the super-speedy U.2 Intel 750 NVMe drive wouldn’t be recognised by our motherboard. This won’t surprise people familiar with the technology as NVMe drives don’t just appear like SATA drives. All kinds of special UEFI BIOS settings need to be used but, of course, there’s virtually nobody to tell you what they are – from Gigabyte or Intel in this case. The best bet you have is forum posts and blogs written by experts and early adopters but the solutions are long and rarely resemble simplicity. Either way, I’ve tried many things, on two motherboards and now I suspect that the drive is broken. But without another drive to hand, you can’t easily switch and check. So that’s why I haven’t had a test rig working properly for at last two months or been able to test a heap of storage devices which can only be properly tested with a fast drive and USB-C connectors. Ugh.
Next up was our Vive test rig. After getting our rig to function with a regular SSD we got an HTC Vive kit in and, well, it didn’t work. None of the troubleshooting worked, HTC struggled to explain things. We established that at least one of the power supplies for the tracking boxes didn’t work but we replaced that. Eventually, another HTC rep who’d been setting up demos at events came round and established that a link box was round backwards. Fixed everything instantly. He said it had been a problem for him at events until he figured it out. It’s not uncommon for problems like this to occur – the ones that make you feel kinda stupid. Later iterations of design tend to make mistakes like this harder to achieve but this problem wasted a couple of days we didn’t have and took two weeks to fix – the lack of expertise, experience and publications on the matter simply weren’t there. That’s one of the reasons I’m writing this.
Finally, today I spent the day setting up various wireless networks to test the latest batch of MU-MIMO routers. One vendor provided a USB dongle which meant we could properly test the MU-MIMO speed boosts on a congested network. After all the issues associated with setting up new networks and running multiple baseline tests we finally got to a point where every time we tested the speedy USB dongle, things actually ran slower. Much slower. All kinds of driver updates and firmware updates have been tried but once again there’s nobody round to ask, there’s nothing online about similar issues (that we can find) and even PR contacting the US-based team couldn't get an answer. Getting a replacement test unit is viable but won't happen until tomorrow and I'm grateful for that. But I wouldn't want to be a regular punter trying to figure this out right now.
Much as I’d like to write a review which says that something doesn’t work and therefore scored terribly, that’s just not fair – you have to know if it’s a general problem or one that’s just down to your review sample (which can often be from an early, production run) not working.
This line almost got blurred a few days ago: I appear to be one of very few reviewers not to gush all over Samsung’s Galaxy Note 7. My reasons are clear on this my results are repeatable and matched my hypothesis. Nonetheless, Samsung is concerned that the unit I have is faulty and I’ll be trying another one just to be doubly sure. I’ll try and film it too in order to be clear on the matter. Who knows, they might be right.
If they are, however, this would also potentially mean that four out of four major new tech products that I've been sent are faulty all at once. What are the odds?
So that’s my rant/whinge/whine out of the way. I only write about such things when several things snowball at the same time. And who knows, someone might be bored enough to read this and have experienced something similar and have an idea on how to fix some of my issues. My usual experience is that the Glorious PC Master Race types demand I turn in my PCMR badge.
But I'm adamant, the more I go through to get stuff working - making use of PR and all kinds of extra journo channels - the more likely it is that you won't have to. So hopefully my pain will be the public’s gain.