From time to time, people contact me to ask me to review their products. That’s how I’m trying out a pair of Mixcder E10 Active Noise Cancelling Bluetooth headphones. So far, they’re a great set of low-to-modestly-priced headphones with excellent product build and sonic characteristics. Here’s what comes with the Mixcder E10 ANC Bluetooth Headphones, from one of their publicity stills:
Clockwise from top left: E10 headphones, old-fashioned dual jack airplane audio adapter, genuine leather zip case, dual mail mini-RCA cable, mini-USB to USB-A cable.
[Click image for full-sized view. Source: Mixcder.com.]
Where Does Ed Get Off, Writing About Audio/Headphones?
I’ve been an audiophile since high school. As soon as I got to college and made some money, I built a HeathKit pre-amp. It went with a Dynaco Tube amp to drive a pair of Arena speakers. My first job after undergrad was as an audio engineer at the Library of Congress (LC). I upgraded to a pair of JBL 4331 Studio monitors. I still used the same pre-amp, with a BGW 250-B solid state amplifier (it still drives my right and left front speakers today).
These days, I run an Outlaw Audio 976 pre-amp/video processor. I also use that same BGW amp, plus 3 Rotel 100-watt solid state amps. They drive a 5.2 setup on Phase Technology speakers. Once upon a time, I attended a summer course on audio engineering at the Eastman School of Music (summer of 1974). This came courtesy of Head Engineer Bob Carneal and the LC.
As part of my job at the LC, I mixed up to 8 channels of audio into stereo. This happened on a Langevin mixing board in the LC’s Coolidge Auditorium. During the “cultural season,” we taped chamber music concerts for public radio distribution. That remote mixing board was backstage. Perforce I had to listen in on a set of “cans” (as we called headphones, back in that now-prehistoric time). My ears hosted Sennheiser headphones that cost over US$300. (The Inflation Calculator says that’s worth $1,737.89 in 2020 dollars.) Those concerts could run two hours, or just a bit over. That explains why I’m (painfully) familiar with extended headphone wear.
E10 Build Quality
As you can see from the preceding photo, $65 to $93 buys a fair bit of stuff. (List price on the website is $112.99; the other prices come from swadeal and Amazon, respectively.) The headphones themselves are substantial, and not overly heavy, at 304 grams (10.72 oz). They’re a closed-ear design. Firm memory foam cups are covered in light leather (or a reasonable fascimile thereof). The headrest is built likewise (memory foam underneath, leather wrapping outside). The cups are labeled “R” and “L” inside. This makes it easy to orient them properly when donning them. The materials are solid and the headset looks able to withstand normal wear-and-tear (and then some, perhaps).
I’d have liked it better if the ‘phones sported a USB-C port for charging rather than mini-USB. But the E10 is what it is. Charging is quick: the first full charge took about 94 minutes. A second full charge only took 60 minutes. The maker claims 60 hours of battery life with ANC turned off. ANC turned on cuts that in half (a still-respectable 30 hours). So far, I’ve not been inclined to push things. I’ve not exhausted the battery after more than 20 hours of active use (with ANC turned on, mostly).
A mini-RCA cable runs the ‘phones in wired mode. Another reviewer claims this produces slightly better audio quality than Bluetooth. Personally, I couldn’t hear any difference between the two forms of input myself. But at age 67 my ears probably don’t work as well as Jupit3r’s (the other reviewer) do, either. Take it under advisement, knowing that a wired connection is handy in signal-rich environments or when flying. I guess that’s when the mini-RCA to dual airplane jack might come in handy. But I haven’t flown on any planes that featured such jacks in quite some time. (I just got back from Washington, DC last Saturday, January 4: we flew a Boeing 737 with USB-A ports.)
E10 Listening Experience and Headphone Comfort
I’ve read numerous reviews (see links at the end of this story) that praise these ‘phones comfort and fit. I’m sorry to say I don’t agree. Though the sound quality is quite good, the phones weren’t comfortable for extended wear. I found myself remembering the “headphone headache” I got in the Coolidge Auditorium after wearing the Sennheiser ‘phones. I couldn’t wear the E10 ‘phones for more than 90-120 minutes without taking a break from their firm grip.
That said, that’s my only gripe with the E10s. The sound quality is good enough that I heard some things amidst my recordings that I’d never heard before. (Not even on my mid-range semi-professional audio rig.) The price is especially attractive, given that you can buy the E10s for under US$100. (In fact, as much as $30 less than that from some sellers.) I have a nearly16-year-old son who’s lost one set of earbuds. He’s also worked his way through two different kinds of Bluetooth headphones over the past couple years. I wouldn’t hesitate for a minute to buy him E10s. I’m enjoying them myself, too. I just can’t handle them for more than two hours at a time. YMMV when you put them on for yourself and see how they fit you. They’re definitely worth a listen, in any case.
Other E10 Reviews Worth Reading
Here is a trio of reviews (including the afore-linked one cited earlier in this story) of the E10 headphones. For more info and different perspectives, check one or more of them out!
KnowTechie: Review: Mixcder E10 — Impressive, lag-free noise canceling headphones
Engineering & Technology: Top-of-Mixcder’s-range noise-cancelling wireless headphones for a less-than-a-ton price.
Head-fi.org: Mixcder E10 Review – Active Noise Cancelling Bluetooth V5.0 Headphones
Here, as promised, is the slide deck for my presentation at SpiceWorld 2019 this morning, about Native boot VHDs in Windows 10. Click this download link for the PowerPoint file (OneDrive).
Native Boot VHDs in Windows 10: Using GPT and Type-2 HyperVisor VHDs (.vhdx)
For GPT, add the “convert GPT” command to the DISKPART script, as follows:
create vdisk file=F:\W10PRO.vhdx maximum=51200 type=expandable
create part primary
When creating a VHD file for native boot, always use MBR partitioning! To upgrade Windows on a native boot VHD, you must temporarily run it as a virtual machine. This means attaching that VHD to a VM.
An MBR partitioned VHD is easy to attach to a VM: mark its Windows partition active. OTOH, a GPT partitioned VHD with only a single partition for Windows requires that you manually create system partitions before it’s usable as a bootable VM (MSR and EFI along with the system partition at a minimum, recovery partition if a complete emulation of “normal Windows 10” is desired).
Say you want to use a GPT partitioned VHD for native boot for some reason (and we can’t think of a valid one). In that case, it is best to first create a Generation 2 VM in Hyper-V, then install Windows 10 on it. This takes care of the partitioning automatically, and does so correctly. That VHD can then be added to the boot menu for native boot, or used as a VM.
I’ve got a couple of nice dual SATA drive docks here at the office. They’re from a Taiwanese job builder named Inatek, model FD-2002. Right now, you can pick one up for under US$30 on Amazon (see preceding link). I’ve used them with great success for drives of all sizes, up to and including 8 TB (Toshiba X300 Performance: ~US$176). Right now, I’ve got one of them populated with 2 4TB HGST Ultrastar drives (~US$95). Until yesterday, I had the 8 TB Toshiba drive paired with an HGST Ultrastar 3TB drive (~US$40, a truly great deal price/performance wise). Then the twist came into play.
Turns out there’s a limit to how much storage will work in one of these low-cost dual drive docks. I hit that limit yesterday!
[Click image for full-sized view. Image Source: Amazon.]
What’s the Twist on Dual SATA Drive Docks?
Turns out that the bigger the drive’s capacity, the hotter it runs. Putting the 8 TB and 3 TB drives in juxtaposition worked fine, until I started using the 8 TB drive heavily. Then, all of a sudden, the 3TB drive started popping in and out of service. Not good! I’m not sure if it was an issue related to the power draw to the drive dock (more than it could handle seems pretty likely to me). OTOH, it could’ve just gotten too hot to keep working. When I took the 3 TB drive out of the dock, it was uncomfortably warm in my hands. I replaced it with a 2.5″ form-factor SSD drive and that combination has been running without a hiccup for almost 24 hours now.
Let this be a gentle warning, dear readers! If you want to pair up storage devices in low-budget dual drive docks, you may want to limit yourself to no more than 10 TB of total storage. Apparently, I hit some kind of wall when I attempted to make steady, serious use of 11 TB (8 + 3) in such a device yesterday. I’m tempted to try it again with a fan blowing on it, to see if lowering operating temps will permit it to keep working. If not, I’ll be more or less convinced it’s a power supply limitation. We’ll see!
I can’t help it. I still love Windows Gadgets. They came in — and went out — with Windows 7, supposedly because of security considerations. I’ve never had, nor heard of such issues in the wild. My old friend and security guru, Deb Shinder, also uses Gadgets. She feels the same away about their demise. Helmut Buhler created 8GadgetPack to bring Gadgets to Windows 8. He’s kept up with them, and they work well with Windows 10. (His current version is v29.0, released in May 2019, so it’s pretty up-to-date.) Each time you upgrade Windows, it sees Gadgets and kills them off. So he’s also written a “Restore Gadgets” routine that recognizes this act, and restores the pre-existing Gadget set-up after each OS-based cancellation. Lately, though, I’ve been posed an interesting Win10 Gadget conundrum.
What’s the Condundrum?
The two following screenshots show the puzzler I’ve been dealing with lately. Notice the display at the bottom of the two CPU Meter Gadget windows that follow. It’s a CPU utilization graph, that shows per-core utilization percentages. The graph on the left-hand-side fills up the entire Gadget panel, while the one on the right-hand-side fills up only about two-thirds. After the most recent Nvidia driver update (which applies to the left-hand item) the Gadget on that PC went from a partial graph area (which applies to the right-hand item) to the full graph area. The PC from which the right-hand item comes is a Surface Pro 3 that uses built-in Intel graphics circuitry that’s built into the i7-4650U processor in that machine. The PC on the left is my production desktop, as the i7-6700 CPU should indicate.
Careful inspection shows the utliization graph at the left fills the full Gadget frame, while the one on the right fills about two-thirds of the same area. What gives?
Here’s the conundrum: I’m pretty sure that there’s something in the nexus between .NET and the GPU driver that’s causing the CPU utilization display area issue. Installing Nvidia’s latest GeForce driver (Version 431.60, which shows a 7/23/2019 release date (but which I just installed today, 8/5/2019) on my production PC fixed the issue. It still persists on all of my systems running Intel on-chip GPUs.
A Minor Niggle Still in Search of a Fix
This is the kind of mystery that guarantees lifetime (or at least ongoing) employment for Windows wizards, gurus, and mavens. I’ll keep messing around with those Intel drivers until I find one that works like it’s supposed to. This has been going on at least since last February or thereabouts, though, so Intel’s obviously under no pressure to catch up with this minor and niggling little GUI detail. It gives me something to do when I get bored, so I can’t complain. And so it goes, here in Windows-World!
PostScript Added August 6, 2019
OK, so I updated my Lenovo T520’s graphics driver last night, too. It includes both an Intel on-chip GPU (HD 3000) and an Nvidia Quadro NVS4200M independent GPU. I updated the Nvidia device to Version 392.56 later yesterday afternoon. This machine had also been subject to the “reduced display” layout shown above for the Surface Pro 3. But after I remoted into that machine this morning, I immediately noticed that the CPU utilization graph at the bottom of the gadget now filled the entire pane, to wit:
This time I loaded Core Temp so that the “not running” error message didn’t appear. Makes the utilization graph a little easier to see and appreciate. So somehow, updating the Nvidia driver (even though I’m not using it) fixed the problem. As a consequence, I’m rethinking my theory of what’s causing the problem. But on my Intel-only systems, the problem persists. I’ll keep working on it . . .
OK, folks. Listen up! Today is the 75th anniversary of D-Day, and the commencement of the invasion at Normandy’s Omaha Beach that would ultimately turn the tides for Allied victory in WWII. General Dwight Eisenhower was informed that casualties could be as high as 75%, but ordered the action regardless. As it turns out, the National D-Day Memorial Foundation determined the number of casualties on the first day of combat at 4,414. Given that 150,000 troops saw action that day, the number could have been much, much higher.
That does not diminish the sacrifice that occurred on June 6, 1944 by one bit. My hat is off to all of those personnel, and to those who were injured or wounded in combat, and to all of their families. My Dad also served in combat in WWII, but in the Pacific theater instead. By the time D-Day rolled around, my Dad was working in the Signal Corps in Manila, in the Philippines, to help restore electrical power to that city. About one month after D-Day, he was seriously burned (and shocked) when a co-worker unintentionally turned on the juice to a 12,000-volt line while my Dad was working on its wiring. He lost most of the skin on his right hand and on the back of his neck, where the current entered (his hand) and exited (his neck). He would be shipped back to the USA, to spend the next four months in recovery and rehab at the famous burn ward at Brooks (then an Army Hospital, now a state-of-the-art Army Medical Center) in San Antonio, TX. One month later, he would meet my Mom, also home from the war for medical treatment, at their mutual hometown of Elizabeth, NJ.
There’s not enough any of us can do or say to recognize and appreciate the sacrifice of all the brave soldiers, men and women, who served our country so faithfully, and so well in this horrible conflict. I’d like to add my thanks and profound appreciation for all of those who served. This goes double for both of my parents, LTC Alfred F. Tittel (USAR, ret, 27 years of active duty) and LT Cecilia K Tittel (nee Kociolek, USAR, honorably discharged in January 1946). Neither of them ever talked very much about their wartime experiences but we could always tell it had touched them deeply, and that they were both honored and grateful to have served their country in time of war and great need.
I’m sitting here at my desk bawling like crazy. As I remember D-Day, this is also a great opportunity to recall and honor my Mom and Dad. Mom passed away on September 11, 2009 and was interred at Arlington National Cemetery in the plot that she will soon share with my Dad. Dad passed away on December 15, 2018 — the same day that Mom was interred at Arlington 9 years before. He’s waiting for his memorial service to join her at Arlington National. Our family is still waiting for Arlington to give us the date for his memorial service, for which Dad requested full military honors, including the caisson and the marching band. The funeral home says it will still be a while yet before we know when that might happen. Whenever it occurs, we’ll be there to honor his passing, and the service that he, my Mom, and all the other soldiers, old and young, also interred there gave this country.
Once again: farewell and thanks to all of them. May you all rest in peace, glory and dignity forever.
The preceding photos are, of course, my Mom and Dad. The photo of Mom was taken in late 1942 or early 1943, just before she was shipped overseas to participate in Operation Torch in Morocco. She would follow the Army through northern Africa (Morocco, Algeria, Tunisia, and Libya), then into Sardinia and Silicy, and onto France before being wounded there in 1944. The photo of my Dad is from the 1990s, in full formal military dress uniform, while attending a gathering of WWII Glider Pilots in Greece, where he was awarded the decoration just under his bowtie for his WWII service.
I got a nice surprise from FedEx last Friday, April 5. An early knock on my door brought a couple of welcome boxes. Courtesy of Jeffrey Witt, Director of Lenovo’s Global Product PR, those boxes housed a couple of 8th Generation i7-based laptops. Mr. Witt graciously acceded to a request I made late in 2018 for a Carbon X1 Extreme, and an X380 of similar vintage. Images and specs will follow, but first, I have another story about some older Lenovo laptops to tell.
A Tale of Two Predecessors
Windows 8 actually has roots back to 2010/2011. By early 2012 Windows 8 was on its way, and technical previews were available. I got hired by Pearson/Que in January to work on a book, Windows 8 in Depth. I also knew Windows 8 was bringing big changes, most notably with the introduction of touch and the tile-based Start Menu. By the end of February, I knew I needed at least one touch-screen unit and one more test machine on which to run Windows 8 Technical Previews. That was the first time I corresponded with Mr. Witt, who was already involved with trade press folks and evals/loaner units. That time around, he helped me get a pretty deep discount on the two machines I bought for my Windows 8 work:
- X220 Tablet: Sandy Bridge i7-2640M Dual Core CPU with Intel HD Graphics 3000, 500 GB spinner, 4 GB RAM, touchscreen, fingerprint scanner. Today, it has a 256 GB Plextor mSATA SSD, 16 GB RAM, and a SATA 256 GB SSD.
- T520 Laptop: Sandy Bridge i7-2640M Dual Core CPU with Intel HD Graphics 3000, Nvidia Quadro NVS 4200M, 500 GB spinner, DVD player, 8 GB RAM, fingerprint scanner. Today, it has a 256 Plextor mSATA SSD, 16 GB RAM, OCX Vertex 4 128 GB SATA SSD, Seagate SATA III 1 TB spinner (replaced the DVD player with a SATA drive module).
I still use both of these machines daily. Considering that they’re now over 7 years old, that’s a real testament to their durability and usability. As I check my Windows Enterprise Desktop blog for Techtarget, I see 41 mentions of the T520, and 58 mentions of the X220 Tablet dating from April 2012 through February 2019. I have really *loved* these machines, and taken them from Windows 7 to 8 to 8.1 all the way to the current build for 1809 (Build 17763.404) on the T520, and the current Insider Preview 1903 (Build 18362.30) on the X 220 Tablet. Until the X1 Carbon Extreme showed up last Friday, the T520 had been my primary “road laptop.” It’s gone on every business trip I’ve been on since I bought it, and many of my personal trips as well. (My Surface Pro 3 and my late, lamented Dell Venue Pro 11 7130 also went along on many, if not most, family vacations.)
Say Hello to the New Kids in Town!
OK then, time for some specs on the new machines. But first, I have to gush about the Carbon X1 Extreme. I think it’s faster and more powerful than my production desktop. My son has completely fallen in love with it and uses it for homework every night. This involves a lot of web surfing, and lots of interaction with various Google platforms elements (Docs, Sheets, and so forth, standard in his school district). He still gets 5-6 hours of battery life out of the machine, which means he can pretty much work untethered. He’s also raved to me about its video playback capabilities. I’ll take his word for it: I’ve had no real time to explore these systems in detail yet:
- X1 Carbon Extreme: Intel 8th Generation i7-8850H (6 cores, 2.6 GHz) with Intel UHD Graphics 630, Nvidia GTX 1050 Ti 4GB, 32 GB RAM, one each Samsung MSLVBxx NVMe 1 TB and 500 GB SSDs, fingerprint scanner and Windows Hello Face recognition, and more. From what I see on Amazon, this machine retails for around $2,600-2,700. It weighs far less than its predecessor, at a svelte 1.84 kg/4 lbs 0.7 oz.
- X380 Yoga: Intel 8th Generation i7-8650U CPU (4 cores. 1.9 GHz) with Intel UHD Graphics 620, 16 GB RAM, Samsung MSLVB10 NVMe 1 TB SSD, fingerprint scanner, and more. Checking Amazon again, I find a typical retail price of around $1,500 for this unit. This one weighs in at 1.6 kg/3 lbs (in round numbers: it’s just a hair under 3 lbs, actually).
It’s really far too early for me to do much more than ooh and aah over the latest laptops in my stable. As I have more time and opportunity to work with them, I’ll be writing about them in more depth. So far, the only thing I miss on the X1 Carbon is an RJ-45 port for direct Ethernet plug-in (but I’ve got a USB 3.0 GbE dongle already anyway, so no biggie; Note added April 9: the unit has a built-in I219-LM GbE NIC–and surprisingly, so does the X380–but you have to buy a special cable to get from its compact proprietary port to an RJ-45. Might be worth acquiring, though). I’m planning to acquire some big (256 GB+) SDXC devices for both machines, to further extend their storage. Another big difference with these newer models is soldered-in RAM (so no memory upgrades). But it’s a tradeoff for their enhanced compactness, I know.
I’ll close with a big THANKS to Mr. Witt and his RTP-based team. I really appreciate the opportunity to work with these machines, and hope to accomplish as many great things with them in the years to come as I’ve done with the T520 and the X220 Tablet. Now, if only i had time to really work them out. . .
Patience has never been one of my leading virtues. Yesterday, while updating my iPhone OS from version 12.1 to 12.2, impatience led me to observe that iPhone Update shows remarkable resilience. Once I’d started the update process over the air. I quickly realized it was going to take a long time to complete the download. At one point, in fact, the time bar said “1 hour remaining.” That’s when I plugged my iPhone into my PC, fired up iTunes, and restarted the update process from there. Indeed the download was MUCH faster to my i7 7700 PC with its near-GbE connection via Spectrum.
All went swimmingly until the machine started its first reboot. It paused forever when the White-on-Black Apple logo appeared on the screen. Figuring that more power to the phone might help speed things up (it didn’t), I disconnected the phone from the PC and plugged it into an nearby iPad charger. Bad move! The update process crashed immediately and I got the “plug back into iTunes, idiot!” screen on the phone shown in the following screenshot:
Egad! I didn’t realize that once you start an update through iTunes, it must run all the way to completion through iTunes. Now I know. Sigh.
Several Interesting Things Happened Next
Thank God for the depth of understanding for human ingenuity/stupidity that the developers who wrote the iOS Update code for Apple baked into the process. The phone’s status screen shows that, even without an OS, it knew enough to show me what to do. On the iTunes side, the program also immediately requested that I reconnect the iPhone via USB so it could attempt repairs. It was smart enough to tell me that it could re-try the update. And if that failed, it could restore the backup which I had fortunately made just before commencing the update process. And if that failed, it said, it could restore my phone to ground zero with a factory reset image. Now THAT’s some quality software engineering.
Of course, now that I know that if I break the USB link between iPhone and PC while an iTunes based update is underway it will then crash, I’ll never do that again. But I was tickled to death to understand that my idiocy wasn’t also rare enough to be unforeseen. I gotta hand it to those Apple update developers. They did a kick-ass job of anticipating what might go wrong during the process, and building in the necessary error-handling smarts to fix things as needed.
All’s Well That Ends Well
I was lucky. The update picked up more or less where it left off and ran to completion. It took a bit longer, overall, than other OTA updates have taken in the past. I neglected to factor in the trade-off between faster download to the PC and USB 2 based file transfer speeds from PC to iPhone. For whatever reason, even though I started out with iTunes through a USB 3 port on my PC, the recovery didn’t work until I plugged the Lightning cable into a USB 2 port. And so it goes here at Chez Tittel, where random bits of lunacy mingle with similar bits of the quotidian. May you learn from my mistake!
Today’s blog post title is more literal than readers may infer. I woke up last night at 12:51 dreaming that my production desktop had been infested. It must’ve been a historical drama of sorts. That’s because it featured a screen takeover scenario. I still remember when you could “catch” viruses and spyware of an interesting sort. They would cause a web browser to spawn an infinitude of Windows. Thus, as I was coming awake, it was from sitting at my keyboard trying to launch an admin command prompt window. It would let me start killing processes and bring the takeover of my desktop to a screeching halt. As Halloween is five days away as I write this, it seems seasonally appropriate in a ghoulish way.
In 2005, when I wrote this book, malware didn’t carry the same risks of financial loss, identity theft, and denial of services that it does today.
Ruminating Over Nightmare On Arbor Drive: Virus Infestation
Back in 2005, I wrote a book for Wiley in its PC Magazine series. It was entitled Fighting Spyware, Viruses and Malware. This title was no best-seller, but it did OK. For me, it was a great excuse to get paid to learn something I needed and wanted to know. In those days, protecting oneself against malware was more straightforward than it is today. The digital world of “cops and robbers” — bad guys trying to foist attacks and steal info of value, good guys doing their best to stymie them — is now more seriously weaponized than in earlier times.
Backup, Backup, Backup
I have a “most important weapon” against malware these days. This is aside from the usual panoply of anti-malware, threat protection, and safe computing practices. I’m talking daily image backups. There’s nothing safer than blowing away the entire preceding and infected or suspect installation, replacing it with a bare metal rebuild. That may not always work against the most insidious of UEFI/BIOS malware or rootkits But it puts paid to everything else. It’s also much faster to restore an image than to clean up after an attack. I’m not inclined to keep a suspect (or infected) image around, either. Macrium Reflect is my weapong of choice. It restores my oldest and slowest PC (a 2013 i5 mobile CPU tablet) in between 15 and 20 minutes. I’ve never been able to get through detection, cleanup and subsequent screening and inspection in under a couple of hours.
“Be careful out there” is as relevant for those who travel the highways and byways of the Internet as it is for the real world. Make sure you’ve got your antimalware act together. Keep a current working backup ready should you need it. Remember also that it’s easier to avoid trouble than to work your way out of it. And for those who observe the holiday, Happy Halloween!
There’s been a lot of complaining about the latest Windows 10 version, 1809, of late. A lot of it appears to be entirely justified, too. But with one exception, my experience with 1809 has been superb. In fact, for the first time I can recall, my post-1809 product PC hits Perfect 10 in Reliability Monitor today. Check this graphic out:
After a few initial hiccups, my reliability on this latest build has been rock-solid. Amazing, in view of history over the past 2-3 years.
Why Say Post-1809 Production PC Hits Perfect 10?
Good question! That latest rise to the top stability index level (10) that Reliability Monitor can report is unprecedented prior to the 1809 install. I would routinely show stability index levels of between 2 and 5. Seldom, if ever, did the value creep much higher. It’s normal for the stability index to reach the top when you don’t use a PC much. (By itself, Windows causes few errors that register in this monitor.) But when you use a machine hard every day — as I do on my production PC, day in and day out — things do go wrong from time to time. And when they do, the stability index usually declines apace.
So this is unexplored and welcome territory for me. I mention it to the world, because I think Win10 1809 may be getting dinged unfairly. Or dinged too much. So here’s a lone but interesting counterexample. It comes from a machine I’ve been using daily since I first put it together back in late 2015/early 2016. Since day 1, the stability index hasn’t gotten above 8 except perhaps once or twice. (It was idle during vacation or business travel.) Upon regular use, it’s seldom climbed about 5, as I said earlier. 15 days into the 1809 install, it’s sitting at a perfect 10 index value. That’s with no errors for the past 10 days. And my usage pattern hasn’t changed at all. My thinking is that 1809 may very well be a better OS than it’s currently believed to be.
But as with all things Windows (10), time will tell!
OK, my long and sometimes odd adventures with Spectre and Meltdown patches are finally concluded. Eight of the nine systems here at Chez Tittel are now patched. That’s as far as I think I’ll ever get because my wife’s PC is built around a Jetway NF9G-QM77 mini-ITX motherboard. Its most current BIOS update is September 2017 from a company for which no word on Spectre/Meltdown updates is available. Thus, for my PCs now eight ninths patched for Spectre Meltdown is as far as I’ll get. It’s been a wild ride. I’d like to document it just a tad to explain what others should be going through, too. Or what they should expect to go through soon.
Steve Gibson’s Inspectre utility finally gives the T520 and its Sandy Bridge CPU a clean (but slow) bill of health.
Getting to PCs Now Seven Eighths Patched for Spectre Meltdown
It all started as we got back from our end-of-year skiing/snowboarding holiday just after New Year’s. Word on these vulnerabilities emerged as soon as January 2. But I didn’t find out until I returned to my desk on January 5. After driving back from the northeastern part of Colorado, I wasn’t ready to deal with a major security flaw. But there it was, and we all had to deal with it. It soon became apparent that Meltdown and Spectre Variant 1 could be handled via OS-level patching (all complete now, thank goodness). However, Spectre v2 required a firmware patch. Or, as it turned out, a series of firmware patches. That’s because the first set for Haswell and Broadwell patches caused as many problems as they were supposed to solve.
The Timeline from Discovery to (Mostly) Mitigated
Here’s a rough timeline for how things unfolded for my PCs, as far as those firmware updates went:
Surface Pro 3 gets a firmware patch 2nd week (1 of 8)
Dell Venue Pro 11 gets a firmware patch late 2nd week (2 of 8)
On 1/15 Intel advises against applying firmware patches
Not much happens with firmware patches
Microsoft issues firmware patch for Skylake, Coffee Lake, Kaby Lake 3/8 (3 of 8)
Dell XPS27 (Haswell) gets a firmware patch 2nd week (4 of 8)
Asrock issues firmware updates for Haswell, Skylake, Coffee Lake, Kaby Lake 3/15 (5&6)
Lenovo issues firmware updates for Haswell, Ivy Bridge and Sandy Bridge 3/15 (7&8)
Hiccups and Lessons Learned
I have an issue with the Dell Venue Pro following its first semi-successful BIOS/UEFI update. It closed the Spectre v2 vulnerability but left the machine unable to reboot normally. I must pop the battery out and remove the power cord before the unit will boot after a shutdown or restart. Thus, I can’t apply the latest update to the UEFI. Among other things, it is supposed to address that very problem. I’m going to have to find and run a flash utility that works from an alternate boot.
That’s what I did with the two Lenovo laptops. Their Lenovo Windows Flash utility works only in Windows XP, Vista, 7 and 8. But I’m running Win10 on those machines. Fortunately, Lenovo also makes the update available in ISO form. It boots to alternate (optical) media and flashes the BIOS from DOS. Even though the Windows utility crashed my Win10 laptops, I eventually booted into DOS to flash them anyway. Along the way, I had to remember to reset boot to support both Legacy and UEFI modes. That’s because DOS is so old, it boots only in legacy mode. On the T520 that was how the machine was set; the X220 Tablet was “UEFI only.” I couldn’t boot to the optical disk until I made that change. Sigh.
One of the Asrock motherboards (Z170 Extreme 7+) delivered the update in a Windows-based flash executable. It was easy to apply. The other, a Z97 Killer Fatal1ty, required using the Instant Flash tool within UEFI. I had to format a USB flash drive to FAT32, unpack the ZIP file to that device, then run the tool from UEFI to apply that update. Took a while, but worked just fine.
No Hiccups Are Nice, Too!
Except for the issue with the Dell Venue Pro and the second UEFI/BIOS update, the Dells and the Surface were by far the easiest to deal with. The Dell Support utility checked for the updates, grabbed them as they became available, and applied them with zero muss and fuss. Ditto for the Surface Pro 3
All in all, while it took longer than I think any of us expected it to, the overall process wasn’t too horrible. Let’s hope this kind of thing doesn’t become too routine, either!