All posts by Ed Tittel

Full-time freelance writer, researcher and occasional expert witness, I specialize in Windows operating systems, information security, markup languages, and Web development tools and environments. I blog for numerous Websites, still write (or revise) the occasional book, and write lots of articles, white papers, tech briefs, and so forth.

Dev Channel 24319 Gotcha

I have to laugh. I’m reading the Dev Channel Build 24319 announcement, and drooling over new features therein.  Especially the USB4 Settings page, which has long been a hobby horse for me. But there’s a Dev Channel 24319 gotcha: I can’t upgrade to a back-rev release on a Canary PC (Build 25314). As I wrote in my March 15 post “Canary Escape Requires Clean Install,” the only way back down the Insider hierarchy is to blow everything away and start over. In a nutshell, that’s the gotcha!

That’s the Dev Channel 24319 Gotcha, All Right

The near-convergence of the build numbers is what got me all excited. The latest Canary channel build is 25314; the latest Dev channel number is 24319. Just one little (leading) digit is what’s getting in the way. The progression looked natural until I paid closer attention and saw that going from 25xxx to 24xxx makes a clean install the only way to get there from here.

I knew this already, but I didn’t really think it all the way through. I got smacked in the face with this harsh reality when I mounted the ISO for 24319 on a Canary machine and got the dialog box shown above early on in the install process. As always when a clean install is the only available option the “Nothing” radio button is the only one available — indeed, it lights up by default, as shown.

Will I, or Won’t I Take the Plunge?

Here’s the deal: I’ve got 35 applications installed on the Canary machine I’m thinking of demoting via clean install. That’s a chunk of work to redo, once the clean install completes. I really want to check out the USB4 capabilities (and indeed I even have a couple of USB4 NVMe enclosure with which to check them). But I probably don’t have time to tackle this for a while. Other deadlines loom larger at the moment.

And boy howdy, is that ever the way things go here in Windows-World sometimes. Can I get another big sigh?

Facebooklinkedin
Facebooklinkedin

SUMo Developer Pays Attention

If you’ve been reading my posts lately, you’ve probably noticed that I’ve complained — just a little — recently about the Software Update Monitor (SUMo) update utility. Over the weekend, the developer himself tweeted me to let me know he’d seen my input and fixed the issue. In fact, he agreed with me that SUMo shouldn’t be recommending a preview/beta version of ANY software. Hence my assertion that the SUMO developer pays attention. He does!

If SUMo Developer Pays Attention, Then What?

He obviously read my recent (Feb 27) post entitled “Update Semantics: Current versus Preview.” And indeed, SUMo is no longer recommending an update to pre-release versions of OneDrive. As you can see in the screencap at the head of this blog post, the current version is indeed recognized as the current version now. It’s highlighted in blue, and comes up with the same version I agree is the current one. Good-oh!

But What About CPU-Z?

I thought I’d caught him out again for asserting this week that CPU-Z needed an update. The download page reports it as version 2.0.5, but SUMo wants version 2.0.5.1. And, sure enough, upon downloading and updating the latest version from the home page, it self-reports as version 2.0.5 (no fourth digit). But after updating and looking at the readme file, here’s what it says:

Sure enough, it really IS version 2.0.5.1. Says so right there!

I’ll be darned. Sometimes the toolmakers know more than the owners/developers do — or what they report, anyway. Very interesting! My thanks to @KCSoftwares: it is nice to know somebody’s paying attention.

Facebooklinkedin
Facebooklinkedin

Intel ARC Drivers Arrive Via WU

There’s a new set of Intel ARC drivers for built-in GPUs (and of course, discrete ARC devices as well). How do I know this? I just updated one of my Canary Channel test machines. During that process, I saw the Intel ARC drivers arrive via WU (Windows Update). Until this morning, I had been obtaining them exclusively from the Intel Driver & Support Assistant.

You can see the information about this latest driver from its Intel download page above. Notice the version number: 31.0.101.4146.

How Do I Know Intel ARC Drivers Arrive Via WU?

Check out the driver version in my Update History from the X12 Hybrid Tablet, captured minutes ago. Compare the version number for the “Intel Corporation – Extension” item and you’ll see it’s identical to the version number from the Intel download page.

ARC Drivers Arrive Via WU.history

The name isn’t terribly helpful, but the version number tells me what I need to know.<\p>

What else I can tell you about this alternate method is that it’s MUCH faster than installing the driver (plus supporting software) from the Intel download page. It took only 20-30 seconds to complete. The full-blown Intel package takes minutes.

Does this mean I will occasionally need to visit the Intel page to update the Intel Graphics Command Center software? Nope. The IGCC that works with Intel GPUs is a Windows Store app. And it updates itself, either through routine checks, or when you try to run that app the next time after installing a new driver.

Hey!  I might actually like this. It’s faster and less work that using the Intel Driver & Support Assistant. Good stuff, and good job: MS & Intel!

Facebooklinkedin
Facebooklinkedin

P16 Mobile Workstation Gets Accolades

Flipping through the Windows news this morning, I was tickled. I saw the very same Lenovo P16 Mobile Workstation sitting next to my desk come up first in MS PowerUser‘s rankings of the best i9 laptops for 2023. That’s right: in this story, the P16 Mobile Workstation gets accolades. And it IS a beast: i9-12950HX, 2 TB NVMe, 128 GB RAM and NVIDIA RTX A5550 GPU.

Here’s Why P16 Mobile Workstation Gets Accolades

It’s got a lot of power. The P16 can tackle tremendous workloads. The cooling is enough to keep it going even when taxed (excellent heatpipes & fans). It’s modestly expandable (my grouse: keyboard must come out to access second M.2 slot for NVMe). It’s especially laden with ports and connections (2 USB-A, 2 USB-C, HDMI, as shown in the lead-in graphic).

There hasn’t been anything I’ve thrown at this machine that it hasn’t chomped its way through faster than any of the other PCs around here. It really, truly is a beast of a machine.

What Might Stop Such a Stellar Purchase?

The list price on this monster is US$9,763.00 as configured. But with current coupons and discounts, it could be yours for a mere US$5,369.65, according to the Lenovo Store. I would have to get special permission from the Chancellor of the Exchequer (my wife, Dina) and be feeling quite flush in the bank account to go forward with such a purchase. But man: what a machine!

It’s got superlative USB4 and Thunderbolt 4support. The P16 delivers superfast IO, internally and through those ports. It’s got a great display, and works like a champ. Given its price tag, you’d have to expect all those things. Totally awesome, though: I completely agree with the MS PowerUser assessment, and understand how and why it showed up first therein.

Who needs a machine like this? It’s not a gaming box. It’s aimed squarely at people who need a portable workstation for developing code, creating media, or handling major, intensive workloads (lots of VMs, AI/ML tasks, 3D modeling, and so forth). Those folks will typically look at the price, understand what the PC can do for their productivity and throughput, and think “Good value for the money.” ‘Nuff said…

Facebooklinkedin
Facebooklinkedin

Winget Suffers Blanche DuBois Effect

There’s a famous line in Tennessee Williams well-known play, A Streetcar Named Desire. It comes from trashy, tragic Blanche DuBois. It reads “I have always depended on the kindness of strangers.” I hope I’m not over-reaching in finding a connection between Blanche and Microsoft’s built-in package manager Winget. Why do I say that Winget suffers Blanche DuBois effect? Because third-party developers must provide package definitions so Winget can handle their updates. Some do, some don’t is my experience on this front.

Overcoming Winget Suffers Blanche DuBois Effect

I turn to other tools to help me catch what happens when the “Blanche DuBois Effect” fails — namely, when a developer or owner does not supply Winget with the necessary package definitions. You can see what I mean by this in the figure below. (Click on it to expand to full-sized view to read what it says for yourself.)

Note that Winget sees 4 items in need of update; SUMo sees 7 or 8.

For the record, Winget sees about half of what KC Software’s Software Update Monitor (aka SUMo) does, to wit:

Tool Count List of items
Winget 4 VSEnt22, SUMo, TeamViewer, Jabra Direct
SUMo 8 Firefox, CPU-Z, Jabra Direct, Edge, Snagit, TeamViewer, WizTree (2)

 

For accuracy, Winget sees one thing that SUMo does not — namely, Visual Studio Enterprise 2022 (abbreviated as VSEnt22 above). OTOH, SUMo sees 5 (or 6) things that Winget does not — specifically, Firefox, CPU-Z (a false positive, in fact), MS Edge, Snagit, and WizTree (in both 64- and 32-bit versions). That’s why I use other application update tools to help me keep up on the ten-plus PCs in residence here at Chez Tittel.

Winget Supplements of Choice Are…

My tools of choice to cover what Winget misses are:

1. Software Update Monitor (aka SUMo) from KC Softwares
2. PatchMyPC from PatchMyPC.com

Why two? Because PatchMyPC sees fewer things than SUMo does. But what the freeware version of PatchMyPC sees, it also updates automatically and easily. The free version of SUMo sees more, but only the paid version tries to update those things for you (and its track record is far enough from stellar on performing updates that I’m not sure it’s worth the US$25-30 you’ll be asked to pay for it).

I’ve learned to use SUMo solely for detection, then I let PatchMyPC handle for me what it can. I do the rest myself manually.

Frankly, though, I think MS should put some of its much-vaunted AI capability to work so that Winget can generate packages for third-party applications on its own with no need to, as Blanch DuBois once did, rely on the kindness of strangers.

Facebooklinkedin
Facebooklinkedin

RAPR V0.11.92 Remains a Real Gem

I’m working on revisions to older stories I’ve written for ComputerWorld. Just yesterday, I revised my CIO story for them about purging duplicate and obsolete drivers from the Windows driver store. For that purpose, there simply is no better tool, nor one easier to use than Driver Store Explorer (aka RAPR or RAPR.exe). Indeed among my many Windows cleanup tools, RAPR v0.11.92 remains a real gem.

Why RAPR V0.11.92 Remains a Real Gem

Here’s the deal: when you update a Windows driver, it gets stashed in a special storage area with all the other drivers. What most people don’t know — including admins — is that when you update a driver, its predecessor remains present. And in fact, it never leaves unless you remove it yourself. In a nutshell: that’s one of the things that RAPR does with ease and grace.

When I wrote the afore-linked CIO story back in 2015, RAPR could help you find and remove duplicate and obsolete drivers. (Note: that item is now carried under the ComputerWorld masthead for IDG’s ineffable reasons.) But you had to do it more or less “by hand.” This took some time and effort to accomplish. No more: now RAPR includes a “Select Old Driver(s)” button. It automatically flags items that might potentially be removed from a target PC’s driver store. Click the Delete Driver(s) button next (see lead-in graphic) and RAPR will remove any selected driver that’s not in actual use.

Why (and When) to Use RAPR

The why comes from reducing the size of the driver store. This applies to any and all windows images for which driver updates get applied. If you put a new one in, RAPR lets you take the old one out. For deployment images — which may run on hundreds to thousands of PCs (or more) — this is especially important.

I’ve gotten in the habit of using this tool monthly. I seldom recover less than 100-200 MB of space. And when GPU drivers come into play (most of them occupy 1.0 -1.2 GB of disk space) those numbers really jump. My biggest-ever savings on an older PC that hadn’t been touched for a couple of years was on the order of 4-5 GB. That’s something fairly substantial.

You owe it to yourself to visit Github and download the latest version of RAPR. Use it to look at your standalone PCs, and the Windows images in your deployment library. I predict space savings all the way around.

Facebooklinkedin
Facebooklinkedin

Canary Escape Requires Clean Install

I kind of knew this already. You probably did, too. But it bears repeating: the general rule for Insider builds is “you can go up by changing your Insider preferences; you can only go down with a clean install.” Simply put, a Canary escape requires clean install.

Thus, this recent Insider email from MS states:

You can only switch to the Dev Channel or other Insider channels that are receiving builds with lower build numbers by doing a clean installation of Windows 11.

Say What? Canary Escape Requires Clean Install

But wait! There’s more. Active Insiders will want to check their email inboxes. In their latest (March 13) email missive “Introducing Canary…” you’ll also find a one-time use code and link through which you can claim a free USB drive. I reproduce the mail-to notification from that web page as the lead-in graphic for this story, in fact. Indeed, I’ve sent my response in to claim my USB drive…

Alas, as you can see in the web page text, delivery time is 6-8 weeks. Does anybody else find this amusing? I keep a whole mini-tray of such drives around for install and repair activities. That’s mostly because Macrium requires users to agree that a rescue disk may only be used on the PC from whence it came. I see 11 on that tray right now (see below). I wonder if the one from MS will fit there, too?

Canary Escape Requires Clean Install.try

Count ’em: 11. There MAY be room for the MS item here, too (depending on size).

Those little flash USBs are Mushkin Atom drives. They’re not the fastest, but they’re compact and eminently usable. As you can see, I keep a lot of them around… And any clean installs I need to do (and, in fact, have already done on my temporarily discommoded X12 Hybrid) will be done and dusted long before that new USB flash drive arrives via mail.

Shout-out to Neowin: Thanks to Taras Buria of Neowin for his March 15 story that brought this to my attention (and got me to read the Insider email all the way to the end… ;-).

Facebooklinkedin
Facebooklinkedin

Port Selection Determines Konyead NVMe Workability

OK, then. I think I’ve figured out what’s going on with my previously reported Konyead mystery. The error reported in that recitation is “The request failed due to a fatal device hardware error.” This happens if I plug into a USB-C port that supports neither USB4 nor Thunderbolt 3 or higher. Thus, it looks like port selection determines Konyead NVMe workability. Interesting!

If Port Selection Determines Konyead NVMe Workability…

I can work around my seeming inability to move the Konyead device from one PC to another by carefully choosing which USB-C port I plug into. I’ve also got an Acasis USB4 NVMe enclosure. It switches back and forth between USB4 and USB3.1 mode without difficulty. The Konyead unit cannot do this, apparently. If it’s presented with a lower-level USB port, it simply refuses to work.

What does this tell me? I think I see the evidence in Device Manager. If you look at the composed screencap at the head of this story, it shows two vital bits of data:
1. It shows that the Thunderbolt Control Center sees the device as Intel USB4.0 (rear layer, top left)
2. It also shows the names of the drivers this connection is using (e.g. WpdFs.dll, WpdUpFltr.sys, and WUDFRd.sys).

When I connect to a down-level USB-C port with the Konyead device, it won’t initialize in Disk Management. It also shows a different set of drivers in Device Manager (Disk.sys, EhStorClass.sys, and partmgr.sys). Those are the same drivers that show up when the Acasis is plugged into the same down-level port. The only difference is, the Acasis device also works with those drivers, too. The Konyead device, however, does not.

The mystery is now somewhat illuminated. I think I’m dealing with the consequences of my experimental idea to “buy the cheapest USB4/TB4 NVMe enclosure” to see what happens. Now I know: it works on the higher-end USB-C ports, but not the lower-end ones. An unforeseen, but at least now visible and understandable, consequence of that perhaps rash approach.

I have to laugh. But indeed, that’s the way things sometimes go, here in Windows-World.

Facebooklinkedin
Facebooklinkedin

HDD versus SSD Widening Price Gap

I’m amazed at the relentless pace of technology growth and change. I can remember paying US$1,000 for a 300MB hard disk (Mac, SCSI) in the mid-1980s. I just saw some ads for NAS drives this morning, and compared them to NVMe SSDs. There’s a serious HDD versus SSD Widening Price Gap going on right now. It’s worth understanding (and watching) — at least, IMO.

Exploring the HDD versus SSD Widening Price Gap

First, let me lay out some price ranges for you. You can buy NAS drives these days in a range from 8TB to 20TB in size. The smaller ones cost under US$200 these days. (The one depicted is on sale at Newegg right now for a measley US$115 or so.) As NAS drive size increases, so does price. A monster 20GB drive will cost you upwards of US$340. (It’s on sale at Newegg, normally US$20 more.) Do the math, and per terabyte pricing falls between US$14.38 (8 TB) and US$18 (20 TB).

Now let’s look at NVMe SSDs. 8 TB is the biggest you can go with NVMe right now (though there are bigger PCIe card drives, I’ll skip them for the nonce). Most 8 TB drives at Amazon fall in a price range from US$1,000 to $1,120. Again, more math produces a per terabyte price range of US$125  to US$140.

The lead-in graphic shows a Seagate Exos 710 8TB NAS drive (below) and a Sabrent Rocket 4 8TB SSD NVMe drive (above). Prices were plucked from Newegg and Amazon this morning. The ratios is what gets me riled up here: on average it goes from 8.69:1 to 7.77: 1.

Price-Performance Pops!

What this all really means is that HDDs still reign supreme for backup and archival purposes where the fact of storage outweighs read/write times. But as somebody who creates daily backups on important PCs who also occasionally has to restore them, I’ll observe that there’s at least a 4:1 speed difference between the two types of media when restoring a backup (sometimes more).

For those with limited patience or time, and especially for those with limited time windows in which to return to full capability, SSDs are increasingly important for necessary backups and storage access.

One more thing: given the ability to put sizable amounts of blazing fast SSD storage as the near storage tier in a multi-tiered storage architecture means that savvy storage buyers can mix and match these two types of storage ever more effectively. That’s how they mostly do things in data centers for cloud and SaaS providers nowadays anyway (except they don’t balk at spending huge amounts on top-tier SSDs like those described in this mind-blowing TechRadar story).

At this bleeding edge of the storage market, it’s clearly a case of “you can’t afford to do this at home” (unless you’re a centi-millionaire or better). But it’s interesting to contemplate how all kinds of drives keep getting bigger, and how storage architects keep figuring out how to deliver ever-huger amounts of data ever more quickly. Again: amazing!

Facebooklinkedin
Facebooklinkedin

Enduring Konyead NVMe USB4 Drive Mystery

Wow! I’m really stumped. I’ve got a Konyead M.2 NVMe drive enclosure that works on only one computer right now. For a long time, I was unable to eject the drive safely. But after backing off the write caching setting for quick removal, and resetting the drive letter from F: to X:, I can now do that. But even so, if I then unplug the drive and plug it into another PC it’s unrecognizable. This enduring Konyead NVMe USB4 drive mystery is driving me nuts!

Showing Enduring Konyead NVMe USB4 Drive Mystery…

When I plug the Konyead into any compatible USB port on another PC (USB3.1 via Type A connector, or USB4 via USB-C connector) it won’t come up. If I go into Disk Management, it immediately throws an error message that says the drive must be initialized. Options offered are MBR and GPT. Choose either one, and the right-hand error box pops up citing a “fatal device hardware error.” Yet, the drive works fine on my Lenovo X1 Extreme (8th gen Intel CPU). What gives?

I’ve tried fixing it with MiniTool Partition Wizard, too. It shows me the device, but also shows it at zero length. Thus, it’s unable to access the raw disk data to find the partitions (and related tables ) that I know are on the drive.

I’ve checked the Crucial SSD’s firmware and driver: both pass the tests from Crucial Storage Executive (the maker’s diagnostic/mgmt tool for this drive). This mystery remains opaque to me. I’m galled that the device works in one PC, but not in others: what’s the point of a removable drive in those circumstances?

Next Steps…

I’ve not been able to find anything about this kind of problem via online searching. I’ll reach out to Crucial’s tech support operation and see if they’ve ever heard of anything like this before. Konyead is impenetrable: konyead.net shows the NVMe enclosure, but all text is in Chinese, and the page for my device won’t come up. They do have a contact page, though, so I suppose I should give it a whirl.

Stay tuned. I won’t quit bulldogging this, but I’m afraid I’m up against what might be an intractable language and culture barrier. We’ll see.

Facebooklinkedin
Facebooklinkedin