Pop Quiz: When Is An Error Message Not An Error Message?

Answer: when it is a complete and utter pack of lies.

Yes, this is about Microsoft. How did you guess?

Back in 2012, after an Apple Time Capsule I’d previously purchased decided to die on me, I got serious about my data and purchased a QNAP TS459 Pro II NAS box. This turned out to be such a wise investment that I subsequently purchased a larger TS-670 and now use that as my “main” network storage device, with the original serving as a “daily backup” server for critical files, with a 7-generation set of backup scripts running at 02:00 every morning…


Except. Any time I try and open any Microsoft Office file that has macros in it, where that file is stored on a NAS, I get a little “warning strip” with a scary-looking message about the dangers of macro viruses. Never used to. This is another one of those Microsoft “improvements”.

Because I can just click the “OK” button and make the problem go away, I’ve ignored it for a while. This might have something to do with my reporting the issue to Microsoft [when I upgraded to Office 2016 for the princely sum of £389] and being told that although they couldn’t offer me the solution for free, I could buy the fix off of them.

Wait, what? No. Just: no.

Being a stubborn so-and-so, I’ve been returning to this problem from time to time, trying to figure out what needs to be done to fix it. This evening I was trawling the web, looking for suggestions, when I found this thread. Buried in there was a post from “philbo2112” [obviously a keen Rush fan] who suggested that the problem could be fixed by right-clicking on Access and selecting “Run as Administrator”.

Well, not quite. However, after doing this and then attempting once more to navigate my way through the Office “Trust Center” and “Trusted Locations”, I discovered that, as “Administrator”, I was no longer able to “see” my network-mapped drives that connect my desktop to my NAS boxes…

Meh. I’ll just use Network Browse and go in that way… And – there was the solution to the problem. Going in “the long way”, revealed that CIFS required the full [canonical?] path name of

\\NAS1\Public\Data\Office Files\

when, of course, Windows Explorer would show the exact same thing as

\\NAS1\Data\Office Files\

The “name of the share” [i.e. “Public”] is “lost” by Windows Explorer once the mapping is made. So… it turns out that all I had to do was hand-hack the path name back to the fill value and that looks to have solved the problem.

Clearly, the programmer[s] who wrote “Windows Explorer” felt they were being helpful by hiding/masking the name of the network share to which my network drive was being mapped. The programmer[s] who wrote the code for Office Trust Center were somewhat more pedantic and wanted the full network path to be explicitly recorded.

If these two products had been written by different vendors, you could sort-of understand how this might have happened. [i.e. because the OS vendor had failed to give clear guidelines on the mapping of network file systems.] But when the programmers for Office and Windows both work for Microsoft, this is inexcusable.


Lies, Damned Lies, and Amazon Parcel Shipping Announcements

On April 2nd this year, I ordered a copy of “Star Wars: The Last Jedi” on Bluray.

At 14:17 on Sunday April 8th, I received notification that it had shipped, via Amazon’s own, in-house courier service.

At noon on Monday April 9th, I checked the delivery status via the Amazon ordering page and was told that the parcel was out for delivery and would be with me by 20:00. I checked again at 19:55 – just 5 minutes before the deadline and was able to confirm that it was still expected.

At 20:05 (and, largely curious to see if stepping past the anticipated delivery date would cause anything to happen with the shipping details) I checked again. I was told that the parcel was now still in transit, but that Amazon were “sorry” that it was delayed and that if it had not arrived by Thursday, I should contact them for assistance. Uh-uh. Where’s my parcel?

So I contacted a representative via a chat window. Top tip – if you ever have to contact Amazon, use a chat window. You will get the chance to receive an email copy of the entire transcript – which is useful evidence.

As to be expected, the loyal Amazon helper fell over himself to apologise – and did really well. To be fair, this was a rather unusual event. However, coming hot on the heels of some flat out lying and deceptive practices from Amazon regarding a printer order, I was inclined to be careful… This was what I got from the Amazon helper:

Please allow me to explain what seems to have happened. We do our best to ensure that all orders are delivered by the date provided when you place your order, but occasionally due to the volume of orders dispatched, there are rare occasions when a carrier receives an order that wasn’t originally assigned to them. However, we still expect your order to arrive as expected.

Be assured, this is not an common occurrence and our transportation team is working hard to eliminate these issues and continually monitor instances like this.”

When I expressed relief that the helper suggested the parcel would arrive on Tuesday April 10th, I got a bit of a surprise:

I don’t want to set any false expectation to your but as the parcel arrive to the carrier facility which was not assign to the order so there might be 1-2 days of delay as we have to ask the carrier to locate your parcel which takes 1-2 days.”

Wait, what? What’s this “locate my parcel” lark? You’ve just spent 10 minutes telling me that the parcel is with the courier and is in the van that delivers to my local area. So there was quite a bit more bluff and bluster about how everything was all in hand, but, basically, that was it.

Or was it? Take a look at the photograph, below, which is the shipping envelope in which it arrived today… See the dispatch date, at the top? “09/04”. Monday. [You might notice I’ve blanked the delivery address, just for the sake of privacy…]

So all that lovely detail about how it had been dispatched on Sunday morning was just complete fabrication. Typical.


Trouble Comes In Threes – Part Four: Microsoft

What, don’t tell me you were only expecting to see three issues within a “Trouble Comes In Threes” montage? Do keep up.

Having more-or-less recovered from the Windows 10 update nightmare, having ordered a replacement printer and having calmed down, yesterday I thought it might be nice to actually grab an hour or two of PC gaming…

You know, what with that being one of the reasons I spent so much money on the thing in the first place…

So there I am, enjoying a decent bit of fun in Skyrim, when my PC performs an instant and full-on crash and gives me a “Blue Screen of Death”.  Now, in fairness to Microsoft, these have become extremely rare since the arrival of Windows 10. This, then, was a big deal. I waited for an error code, but got,

WHEA Unrecoverable Error

In other words, nothing useful. The system rebooted and I had just re-loaded Skyrim from the most recent save when… another blue screen. Same error code.


Nothing for it but to wait, reboot, and try again. Whilst obviously checking the logs files and looking for obvious problems – nothing on show…

When the computer crashed for a third time, the Blue Screen Of Death gave me a “new” error code – certainly not one I’ve seen before. This time I got,

Clock Watchdog Timeout

Interesting. And that was it…

There’s nothing in the “Update” logs to suggest that anything has changed in the build, but then Microsoft aren’t exactly truthful. Or it could be that they had reached out and turned on a piece of their spyware that isn’t normally active and that this occasionally-run piece of code was clashing with the gaming environment. That’s the problem of Windows 10. You just can’t know what it’s doing. When it does tell you anything [which is a pretty rare event these days] then the information is minimal. Even Event Log has become largely useless.

Is this it?

The end of the shambles that has been the latest Windows 10 Update? who can tell? This is part of the charm of Windows 10 – you just don’t know what is going to screw up next – or when.



Trouble Comes In Threes – Part Two: Hewlett-Packard

After the complete disaster that was Microsoft’s KB4088776 update, one of the things I had to do was to re-install the printer drivers for my two printers. I have an Epson XP-950, with I use almost exclusively for photo printing, along with an HP OfficeJet Pro 8100, an A4 duplex (double-sided) workhorse printer. The HP has been absolutely fantastic – it’s compact, quiet, extremely fast and economical to run, even with the very expensive HP inks [I only use original inks].

After the rebuild of my “Office” system, I downloaded and installed the latest HP printer driver, but, in a momentary lapse, allowed it to perform a “complete” installation, so in addition to the basic drivers, I also ended up with all the “phone home” technology.

Later that day, I went to print off some documentation I had been working on, only to have this HP printer throw an error. “Print head failure…” was being reported… Wait, what? I had just been using the printer earlier that same day, from Mint Linux running from my little fanless system, and all had been perfect. What gives? OK, print head…

The 8100 comes with a removable print head [it used to be possible to purchase spares] and as a result I lifted out the unit to give it a thorough inspection. There was absolutely nothing showing from a physical point of view, even scanning the print nozzles using my camera and a macro lens…

I took an hour to painstakingly thoroughly clean everything with isopropyl alcohol and put it all back together, and… a different error. Reboot. A different error. Reboot. A fourth error. In short, the printer was completely toast. According to my records, I purchased the 8100 from GA Business on July 20th, 2014 – so I’d had it roughly three and a half years. In all that time I had probably printed 500 sheets, half of which would have printed double-sided, so say 750 sheets of paper… Print quality was flawless, right up to the moment it died.

In fact, the only thing that was remotely unusual was that the failure happened the moment the Windows “HP software” connected to the Internet and phoned home.


What do you think?


Trouble Comes In Threes – Part One: Microsoft

Back on the 13th or 14th of March, Microsoft rolled out a massive update to the Windows 10 OS. Like all Windows 10 updates, it’s invisible: unless you check your network traffic, there’s no evidence it is even being downloaded.

There is a long and painful history of what went wrong here (Microsoft Support Site): the short story is:-

KB4088776 completely trashed two builds.

Build One – Symptoms – “Dead” Monitor

A machine running Windows 10 was left “alone” during a copy of files to external disk for backup purposes. On return only 2 out of 3 monitors were working. All three screens are identical Dell U2415W monitors (running a Dell Driver date-stamped December 17, 2017), connected to the system via DVI-D, HDMI and DisplayPort, respectively. The “DisplayPort” monitor was dead. When I checked via my nVidia Control Panel software, it reported that one of the three monitors was no longer recognised – the panel looked like this:-

Build One – Fix

Using the Windows Installation USB Key provided when I purchased the Windows 10 License [64-Bit Pro edition] I was able to boot the system as far as the first setup screen, from which I was then able to select “Repair” and this allowed me to “roll back” the KB4088776 update. There had been no other changes to the system in that time.

Conducting the restore didn’t solve the problem 100% – I was still left with only two monitors, but I was now able to go in and reactivate the “de-selected” monitor and get back to my triple-screen setup. Eventually I was left with something that looked like this:-

Build Two – Symptoms

Unfortunately, things got worse, a lot worse, from there. The hardware that I use for gaming includes an amazing little unit from Akasa that slots into a 5.25” drive bay and gives me no less than four 2.5” drive slots. It looks like this:-

Because the drive bays make it easy to replace the drives they contain, I can swap between different Operating Systems in the time it takes me to power down, swap drives and power up again… So I bought a pair of Windows 10 Pro 64-Bit licenses [£200 each – ouch!] and away I went…

On the second drive, I used the same method to boot from the Windows 10 Installer USB stick, but when I attempted to “Repair” the OS, I was told that there was nothing there to be repaired… What? I went back to the now-working “Games” build and booted it, then used an eSata dock to take a look at the C: drive from the “dead” Office build…

What I found was a completely trashed operating system. The update process had left two Windows directories and had also dumped some error files in the “service” partition at the front of the disk. The above-linked series of posts on the Microsoft site provide details. In short, the OS was toast.

Build Two – Fix

Nothing for it: a wipe of both the C: and D: volumes [after making careful note of what had been installed] and a clean build of the OS. For the most part this went smoothly [it took all day] with the exception of the re-installation of Nuance Software’s OmniPage 18.

That refused to go back, claiming I had exceeded my license count… I dropped a line to their support desk and eventually I was granted an extra “activation” to get the job done – after being warned that version 18 was not compatible with Windows 10. I can’t find a more recent version, so: whatever.


I don’t know what Microsoft are doing with Windows 10 – other than completely and utterly screwing it up.