Archive for PEBKAC

Unexpectedly weak

I have fairly decent Wi-Fi at the palatial estate at Surlywood. Then again, it’s a fairly small perimeter. A house that takes up a whole block — let’s say, oh, the 1600 block of Pennsylvania Avenue Northwest in Washington — well, that’s a different story:

Barack Obama moaned about the White House’s Wi-Fi coverage in an interview with CBS’s Super Bowl pre-game show. “This is an old building so there’s a lot of dead spots where the Wi-Fi doesn’t work … no, actually it’s an issue,” Obama explains.

See? You thought the President was going out of town to go loaf somewhere and/or play golf. He’s actually looking for a better signal.

Michelle Obama agreed, adding that their two daughters, Sasha and Malia, get “frustrated sometimes.” Obama hopes to fix the “whole tech thing” for “the next group of folks” who will move in.

I just hope there aren’t any mail servers in the bathrooms.

(Via Fark.)

Comments




Trojan apparently slain by geek

Not even malware is hackproof, it appears:

Users tricked by spam messages to open malicious Word documents that distribute the Dridex online banking Trojan might have a surprise: they’ll get a free antivirus program instead.

That’s because an unknown person — possibly a white hat hacker — gained access to some of the servers that cybercriminals use to distribute the Dridex Trojan and replaced it with an installer for Avira Free Antivirus.

Good thing, right? But still against the law:

Although replacing known malware with an antivirus isn’t an activity most people would consider a hacking crime, it’s likely against the law in most countries. A whitehat hacker who figured out a way to penetrate Dridex servers and tamper with the malware distribution channel may have done so discreetly to prevent being detained or prosecuted by law enforcement authorities.

And of course there’s a worst-case scenario:

A competing theory is that Dridex operators intentionally included the AV installer, possibly to throw off the detection process of other AV engines.

Which might be plausible, since the installer does not actually autorun: the person receiving it has to run it manually.

(Via Fark, with the kind assistance of @SwiftOnSecurity.)

Comments (1)




Obviously not too damn much

I suspect this incident will never make it into a future edition of How to Make Friends and Influence People:

Okay, maybe in the Appendix, under “Bad Examples.”

Comments




Poor but dishonest

Every barrel, it seems, contains a few bad apples:

GRACE Marketplace thinks of itself as being the Walmart of homeless centers.

In one centralized location in Gainesville, Florida, it offers end to end services: substance abuse counseling, help with signing up for public benefits such as food stamps, showers, restrooms, meals, a place to store personal belongings, an adjacent tent village called Dignity Village, and more.

Unfortunately, it just lost one crucial service: namely, the free Wi-Fi that could have helped Dignity Village residents to find or apply for jobs.

And how did this happen? Pretty much the same way a lot of people with roofs over their heads lose their service:

“We would love to be able to provide Wi-Fi out here, but we don’t have any IT support,” said Jonathan DeCarmine, GRACE Marketplace operations director. “We were notified by our Internet service provider that there were people downloading things illegally, and if we didn’t put an end to that, they would turn off Internet to the entire property, which would keep us from being able to do business and provide services.”

Meanwhile, at the next level up:

Theresa Lowe, executive director of the North Central Florida Coalition for the Homeless and Hungry, said she has no plans to turn the Wi-Fi back on. They had some security restrictions in place already, but people found ways around them. She said there can be hefty fees for illegal downloads, and that’s something the center can’t afford.

“We had a couple complaints from our provider and notified everyone, ‘please don’t do this, we’ll end up losing the service,” and it happened again, so our decision was to disable the Wi-Fi because we would be charged,” Lowe said.

Those whose business model depends on depicting the homeless as saintly and utterly without blame will be crying into their kale smoothies; as with any other community, any other demographic, “good” and “not so good” live cheek by jowl.

Comments (2)




A new crease in the black hat

Anyone who owns a domain has likely received a “bill” from a third party offering to renew that domain at some ridiculous multiple of the actual registration price. Enough people have caught on to this scheme that now the scamsters are having to pretend they’re offering a service:

SEO pitch for wendex.net

Obviously the most important thing here is “SECURE ONLINE PAYMENT.” Amount of said payment: $63.00.

In the fine print down below:

You have received this message because you elected to receive special notification proposal. If you no longer wish to receive our notifications, please unsubscribe here or mail us a written request to US Main Office: SEO Domain Registration Company, Los Angeles, CA 90036, Email: seodomainregservice@mail.com or Asia Main Office: SEO Domain Registration Company, Shenzhen Futian, Email: seodomainregservice@mail.com. If you have multiple accounts with us, you must opt out for each one individually in order to stop receiving notifications notices. We are a search engine optimization company. We do not directly register or renew domain names. We are selling traffic generator software tools. This message is CAN-SPAM compliant. THIS IS NOT A BILL. THIS IS A NOTIFICATION PROPOSAL. YOU ARE UNDER NO OBLIGATION TO PAY THE AMOUNT STATED UNLESS YOU ACCEPT THIS NOTIFICATION PROPOSAL. This message, which contains promotional material strictly along the guidelines of the CAN-SPAM act of 2003. We have clearly mentioned the source mail-id of this email, also clearly mentioned our subject lines and they are in no way misleading. Please do not reply to this email, as we are not able to respond to messages sent to this address.

I want to see how a “written request” gets to the SEO Domain Registration Company without a street address in Los Angeles, CA 90036 (near Hancock Park and the Miracle Mile) or however the Chinese sort these things out in Futian district, Shenzhen.

Comments (2)




Marked for death by Information Services (12)

All CAPTCHAs are annoying, but some are more annoying than others, and the one with the four-by-three grid of thumbnails — “Select all photos containing X” — currently sets the curve for Maximum Farking Irritation, especially if you have to go through several cycles to persuade it to shut the fark up.

Comments (3)




Well, they do own the name

This was trending on Facebook — at least, the Facebook they send me — last night:

Facebook screeshot: iPhone: Apple May Release New Smartphone Named After iPhone 5, Report Says

Um, say what?

Mark Gurman of 9to5Mac reported Friday that the 4-inch “iPhone 5se” would replace the iPhone 5s, with features including an 8-megapixel camera, an A8 processor and Live Photos.

Oh. Why couldn’t you have said “Apple May Release An Updated iPhone 5”?

Update: Someone wised up. It now says “New Smartphone Model to Replace 5S.”

Comments (1)




More arbitrary levels

Bill Peschel had to go rooting, so to speak, around his Web site:

I had received an email from HostGator telling me that they couldn’t back up PlanetPeschel because of “inode,” which was about as unhelpful a message as one could get.

Fortunately, HostGator had a page explaining what the heck inode means. In plain English, I have too many files on my shared hosting. More than a 100,000 of them. They don’t mind my having them, they said, they’re just not going to back up all of them. And if you get to 250,000, they warned me, I’ll have to pay them and make some changes.

(Shared hosting, if I remember right, means my website’s on a server with a bunch of other sites.)

This is indeed what shared hosting is. I have it. I have five sites on that machine, but I’m not alone.

Anyway, this surprised me. I’ve got a lot of files on Planetpeschel, but 100K?

Depending on what the cache has done lately, I have somewhere around 25,000 files.

I have a different backup issue. On my WordPress installs, a plugin called wp-db-backup gzips up the database and emails me a copy once a week — except here. The issue is that this database is freaking huge: on the order of 75 megabytes. Doesn’t sound like a lot for nearly 20,000 posts and about two and a half comments per, but the catch is that gzip brings it down to 21 or 22 MB — and the host’s email facility won’t handle anything in excess of 20 MB. Workaround: while I’m not working on the site, I set the backup to run manually, and drop the resulting gzipped file on my desktop.

Bill’s excess-files problem, however, isn’t related to this at all:

I have a plug-in on WordPress called Autoptimize, which saves bits and pieces of the site to call them up quickly. Turns out my settings told it to save a lot of files. Like, 65,000 of them.

Once I got rid of them, HostGator thanked me and said they’ll back up my site next week. Everyone’s happy (except me; I had to put a note down on my calendar a week from now to check my inodes and see if they’re swelling again).

I hadn’t heard of this plugin, so I went looking:

Autoptimize makes optimizing your site really easy. It concatenates all scripts and styles, minifies and compresses them, adds expires headers, caches them, and moves styles to the page head and can move scripts to the footer. It also minifies the HTML code itself, making your page really lightweight. There are advanced options and an extensive API available to enable you to tailor Autoptimize to each and every site’s specific needs.

Which sounds nice. Then there’s this:

If you consider performance important, you really should use a caching-plugin such as e.g. WP Super Cache or HyperCache to complement Autoptimize.

Hmmm. I went straight to the caching program, myself. It saves bigger pages — full-fledged HTML static pages — but a lot fewer of them.

Comments (2)




8080 in ’80

Back in the day, few of us could aspire to these heady heights of technology, but oh, how we dreamed:

Advertisement for IMSAI Computer System

Information Management Sciences Associates, Incorporated, founded in 1973, sort of still exists today: Fischer-Freitas, which was actually set up by two former employees to acquire IMSAI after its 1979 bankruptcy, continues to provide parts and support for these Ur-machines.

(“Optional 16-bit”? 8085, maybe?)

Comments (2)




Bandwidthout

It is de rigueur for us occasionally peripatetic types to grumble about the Wi-Fi offered in this hotel or that resort. (Me, I hunt around for a Cat5 jack, and usually I find one. Yes, I carry cables.) But there’s another side to that story:

I have way more sympathy for hotels and their wifi systems. We installed a wifi system in a 100 site campground in Alabama. That system has become a data black hole — no matter how much bandwidth I invest in, people use more. Every night it seems like there are 300 people on 100 campsites all trying to stream a movie in HD. I am not sure it will ever [be] enough, and we get no end of speed complaints despite having an absurd T1 bandwidth into the system. I can’t see myself ever investing in such a system again.

Not that a T1 is all that doggone fast: 1.544 Mbit/sec. Divide that by 300 and you have approximately AOL circa 1994. Still, the principle seems clear: the more bandwidth you have, the more gets used. Traffic planners, in the automotive sense, have known this for years, and have occasionally behaved accordingly.

Comments (3)




The matrix rebloated

I’m not the guy who writes the code, but I’m usually the guy who has to answer the halts, and few things on a green screen are quite as frustrating as this: An array index is out of range. There follow several, usually four, options, none of them good. I lean towards D, which is basically C (cancel program) plus spool a memory dump.

It happens to the best of us:

Applications written around arrays have caused much destruction. I once had a matrix mechanics job, a twelve-hour program that ran on a supercomputer, fail in the ninth hour because of an array problem. The great continent-wide communications crash of a decade ago was caused by a mis-defined array. Two major stock market recording debacles occurred because an array was undersized — the same array in both cases, ironically enough.

Then again, a lot of this is legacy code that we don’t have time to rewrite. I swear, there are still bits of junk from the 1990s being called here in 2016.

This is probably not the time to note that we have one office subsystem running on a Windows XP box.

Comments




Hiding behind the cloud

I have long suspected that some of the alleged silver linings of cloud storage were actually nothing more than zinc, and this doesn’t make me feel any better:

In late 2012, I decided that it was time for my last remaining music CDs to go. Between MacBook Airs and the just-introduced MacBook Pro with Retina Display, ours had suddenly become a CD-player-free household.

The 150-or-so CDs in question were living a second life as AAC files in my iTunes library, but a niggling thought persisted: what if something better than AAC came along? What if I wanted a higher bitrate after all? What if?

The solution seemed obvious, since commodity-level multi-terabyte drives weren’t ubiquitous then: send it up to the cloud, and specifically the Amazon Glacier segment thereof. Sixty gigs for less than a buck a month? Sounds good to me.

Then those drives materialized, and why spend even a buck for storage anymore? So he decided to retrieve the files, and this happened:

Glacier data retrievals are priced based on the peak hourly retrieval capacity used within a calendar month. You implicitly and retroactively “provision” this capacity for the entire month by submitting retrieval requests. My single 60GB restore determined my data retrieval capacity, and hence price, for the month of January, with the following logic:

  • 60.8GB retrieved over 4 hours = a peak retrieval rate of 15.2GB per hour
  • 15.2GB/hour at $0.011/GB over the 744 hours in January = $124.40
  • Add 24% VAT for the total of $154.25.

Plus bandwidth costs, bringing the bill to $158.83, which would buy several terabytes of drive.

More and more, we expect cloud infrastructure to behave like an utility. And like with utilities, even though we might not always know how the prices are determined, we expect to understand the billing model we are charged under. Armed with that understanding, we can make informed decisions about the level of due diligence appropriate in a specific situation.

The danger is when we think we understand a model, but in reality don’t.

Yep. I don’t think I’d have figured that out from the Glacier pricing FAQ.

Disclosure: I am a customer of Amazon Web Services, though I have never used the Glacier service.

Comments (3)




Live by the link, die by the link

I should have left this tissue of organic fertilizer in the spam bin. It was titled “A few recommendations that can help both of us,” which inevitably means it will help the sender and may or may not do the recipient any good. Decide for yourself:

Webmaster,

We are writing to alert you to the presence of harmful backlinks on your website. These links, that direct to [domain name redacted], were placed on your site by our former SEO management company. That company willfully violated Google’s Terms of Service, which resulted in a penalty being levied against our company. In order to remove this penalty, we must ask your assistance. Please delete the known backlinks to [domain name redacted], hosted on your site at:

http://dustbury.com/archives/cat_blogorrhea.html

Your compliance with this request is greatly appreciate. Have a nice day.

Generally, anyone who uses the word “backlink” unironically can be assumed to be a scoundrel or a fool.

What’s hilarious about this is the origin of said, um, links: this domain, once upon a time, belonged to a blog which once — well, twice, actually — hosted the Carnival of the Vanities, and I always linked to the Carnival host as a matter of historical reference, since I contributed the first piece to the very first Carnival, way back when. There’s now a storefront sitting there, and their current SEO management company apparently got its BVDs horribly knotted at the thought of an incoming link that would not sell any product.

How much do these links harm me? I have a better chance of winning the freaking Powerball. Just the same, I took them out, on the basis that I don’t want to hear from these whimpering sons of bitches ever again.

Comments (1)




And lo, there was debugging

A cry from my techie side (which is actually barely more than a corner) from last month:

After installing WordPress 4.4, I encountered a minor anomaly. Short version: In posts included in two or more categories, the categories are now listed in the post heading, not in strict alphabetical order as they used to be, but in the order of their assigned ID numbers, whatever they may be. I left a note at the support forum, indicating what I thought might be the issue, and expressing some nominal amount of dismay.

Unusually, no one at the forum deigned to respond to my dismay, but the problem was quietly fixed in 4.4.1, possibly as a by-product from this bug.

Comments (2)




From the As If files

This character has big dreams and, one suspects, no way to make them come true:

What type of Web hosting plan would be good for a site that gets 100k visitors a day. Would a shared hosting plan be able to handle this much traffic or vpn or cloud hosting? Not very tech savvy, thank you.

A guy who evidently has never had a site of his own before expects a hundred thousand visitors a day. This is like trying to run HuffPo off Weebly.

I imagine the surfer dudes who host this site would gently, and then not so gently, suggest I move up from the $10-a-month plan if I started getting 100k visitors a day. (The most I’ve ever had on a single day was 13,636, on 12 May 2009, and at least 11,000 of those came from Instapundit.)

Comments




Upgrade or die

This is the last week for Microsoft Internet Explorer 8, 9, and 10:

Microsoft is ending support for Internet Explorer 8, 9, and 10 next week on January 12th, releasing a final patch encouraging users to upgrade to one of the company’s more recent browsers. The end of support means that these older versions of Internet Explorer will no longer receive security updates or technical support, making anyone who uses them much more vulnerable to hackers. A recently-announced patch will deliver the last few bug fixes, as well as an “End of Life” notification telling users to upgrade to IE 11 or Microsoft Edge — the company’s successor to Internet Explorer, built for Windows 10.

When I heard about this, well, you can imagine my response:

NetMarketShare reports for December that version 11 now dominates among IE users, with 25 percent of the total browser market. Curiously, IE 8 is next (just under 9 percent), followed by 9 and 10 — but 6 and 7 still show fractions of a percent.

My own stats package reveals similar numbers, plus one startling statistic: out of 3496 IE connections in the last ten weeks, two were through version five, which Microsoft is supposed to have killed off at least a decade ago.

Comments




You’ve got (entirely too much) mail

I don’t do Gmail, but I figure any client for any email provider can do this to you:

I’m looking at my G-mail. I have over 10,000 messages there. Why? Because I thought maybe I should save one, and yes, maybe I should save that one too, and pretty soon I was saving everything and now I’ve got 10,000 messages. Theoretically, there might be some important information in there somewhere, and occasionally I have managed to retrieve some useful bit, but is it worth carrying all this baggage around for that one, possibly useful, bit of information that I dredge up every six months or so? If I was having to pay for this storage, no it wouldn’t. But I am willing to carry the mental load of thinking there are useful things stored safely away in this pile of verbiage. So now I am thinking that maybe I should just delete everything and start over.

I threw away around 8000 messages at the end of the year. I still seem to have, um, 41,287 in the archive. Then again, the archive goes back to 1997. And spam is not a consideration in this count: except for items mocked in this space, spam is deleted more or less upon receipt.

Comments (9)




You should be so lucky

Potential for disillusionment: high. In fact, make that very high:

Yahoo Answers screenshot: If I start learning programming at the age of 19, when will I become rich?

He thinks he has a plan:

I will be attending UCR next year after I transfer from the community college I am currently attending. I plan on graduating and having my bachelor’s degree by the age of 23. If I work on the courses counted towards my computer science degree and study other programming languages online through sites like W3Schools.com. What will I be capable of? Could I make an app and become rich before or at the time I turn 30?

Wait until he finds out that “rich,” in programmerspeak, means “not living on cat food.”

Comments (4)




Every existence has a bane

This isn’t mine, but it’s awfully damned close:

Yahoo Answers screenshot: How can I create a free website with domain name?

We have produced an entire subculture of people who want things of value and who don’t want to pay for them. I really think the Prince of Darkness (curses be unto him) is going to have to build on an addition to hell to make room for them all.

Yeah, I hear you: “It’s probably just a kid.” Kids with no sense of propriety or property grow up to be adults with no sense of propriety or property. There aren’t enough roads to Damascus for all of them to wake up in time.

Comments (5)




And apparently nowhere to go

Just the same, it looks like you actually can get there from here:

Now I’m curious to see their printed schedule.

Comments (1)




We’re so confused

And having had similar troubles myself, I can relate:

What? Don’t look at me. I never know my Wi-Fi password.

(Via HelloGiggles.)

Comments (1)




Take one tablet

One week and one day with one low-end tablet. How does it work? Pretty well, actually, but it’s not exactly the answer to anyone’s prayers.

Okay, maybe the prayers of Jeff Bezos.

Comments (3)




Program exit

I’m hoping this man had an orderly shutdown:

William Ralph “Bill” Fink, 46, of Belleville, Ill., born July 28, 1969, in Belleville, Ill., encountered an unhandled exception in his core operating system, which prematurely triggered a critical “STOP” condition on Wednesday, Dec. 16, 2015.

Bill was an avid technophile, program developer, and educator, whose master functions were harnessed by Microsoft Corp. as a technical evangelist. Some of Bill’s most impactful component subroutines centered around video games, coaching youth sports, building elaborate displays for Halloween, and spending time with family.

And because you need to know these things:

Diagnostics indicated multiple cascading hardware failures as the root problem. Though his hardware has been decommissioned, Bill’s application has been migrated to the Cloud and has been repurposed to run in a virtual machine on an infinite loop.

(Via Matt Prichard.)

Comments




They’re keeping close-mouthed about it

Perhaps no one anywhere is immune to the possibility of identity theft:

A database for sanriotown.com, the official online community for Hello Kitty and other Sanrio characters, has been discovered online by researcher Chris Vickery. The database houses 3.3 million accounts and has ties to a number of other Hello Kitty portals.

The records exposed include first and last names, birthday (encoded, but easily reversible Vickery said), gender, country of origin, email addresses, unsalted SHA-1 password hashes, password hint questions, their corresponding answers, and other data points that appear to be website related.

The earliest logged exposure — the first time anyone accessed this data in the wild in a manner consistent with leaving log entries — was the 22nd of November.

Sanrio, as well as the ISP being used to host the database itself, have all been notified. An automated email from the ISP confirmed that the incident notification was logged, but no further details are available.

(Via @SwiftOnSecurity.)

Comments (1)




Twenty-five and counting

Today, it appears, is the 25th anniversary of the World Wide Web:

If the web were a person, it wouldn’t have trouble renting a car from now on: the world’s first website, Tim Berners-Lee’s World Wide Web, went online 25 years ago today. The inaugural page wasn’t truly public when it went live at CERN on December 20th, 1990 (that wouldn’t happen until August 1991), and it wasn’t much more than an explanation of how the hypertext-based project worked. However, it’s safe to say that this plain page laid the groundwork for much of the internet as you know it — even now, you probably know one or two people who still think the web is the internet.

More than one or two. I blame Microsoft, which used to call its Web browser “Internet Explorer.”

It still stuns me a little to think that I’ve had an outpost on the Web for most of its existence. But it’s true: this little site went live on the 9th of April 1996, and has had some form of update every single day since the summer of 2000. Eventually, I suppose, the world will move on to something else. Then again, so must I, and so must we all.

Comments




I take it back

This seems improbable:

From their FAQ, a hint as to how it might work:

Does ‘unSend.it’ actually remove the entire email?

unSend.it removes all content from the body of the email (including any attachments) you sent. The email itself will always remain in your recipient’s inbox along with the subject line — however, all email message content will be removed.

This suggests to me that it’s sending a second, edited copy of the email, duplicating as much of the original header as possible without running afoul of mail protocols, and overwriting the old with the new.

Which also suggests an issue: Suppose your recipient is the sort of person who stashes incoming mail into a different folder once it’s read. Can the service tell where that message has been relocated?

Comments (1)




We must first acquaint ourselves with device

In recent years, 42nd and Treadmill has been working on minimizing the cost of the annual Christmas party while not actually discouraging people from coming. It’s been brought in-house, some decent caterers have been located, and they’ve been giving away stuff like a day off with pay (retail value $64 and up). But this year, they spent $300ish on door prizes, and I know because I got one of them: specifically, one of six bottom-of-the-line Amazon Fire tablets, currently selling for $49 each and out of stock until the Monday after Christmas.

Two things happened that I expected: first, the little gizmo wanted to download a crapload of updates, and then it took Windows 7 three tries to install a device driver for it. One thing I should have expected, but didn’t: it’s almost pointless to own one of these critters unless you’re also an Amazon Prime member. A thirty-day trial is underway, but I was considering going that way eventually.

Oddly, I was considering buying a Fire anyway; I’d gotten to the point of cross-shopping the various models, and had actually dropped the next one up into my cart, withdrawing it once I got hit with storm expenses. And the next one up is decidedly richer in function. But since I am the rankest of amateurs at this sort of thing, the smallest Fire will do for now. And to my delight, this here Web site looks pretty darn good on it. (Take that, you “responsive” theme-hawkers!)

Comments (4)




Chrome for the holidays

During the power outage a few days back, when I fled to a hotel to avoid freezing my keister off, I hauled Toshi the Road Warrior, my ancient (turn of the century) laptop, along with me. It still works, but of necessity it’s slow: it’s a 1.1 GHz Celeron with a system-board maximum 512 MB of RAM. On Windows XP, of course. I mentioned this situation to the sysadmin; he offered to lend me a Chromebook to play with, which I picked up yesterday.

So I plugged in the little metal slab, from Samsung — this is it — and realized that I had no idea what my wireless password was. (Yes, it’s really been that long.) Fortunately, I did know how to get into the router, and I reset all the wireless parameters. (Which meant that I had to bring up the laptop and update it with the new wireless stuff, in case I have any ideas about using it again.) It’s a nice little machine, I suppose, but it’s going to take some getting used to. First order of business is to get a proper mouse. I can work the Chromebook’s touchpad — it’s not so different from the one on the laptop — but I don’t like it either.

Comments (4)




Do not turn off

Most of the time, Windows accumulates updates, and then when you’re ready to shut down — usually on a Tuesday night after a West Coast basketball game — it flashes the Evil Shield at you, letting you know that the machine’s going to be tied up for another 15 minutes (my home box) to two hours (my work box). But sometimes it insists on rebooting the moment the update has arrived:

Most updates no longer create an unusual machine state that requires a reboot to resolve. There are still a few that do. In an ordinary consumer environment, the remaining problem is small enough to be ignored (or at least small enough that there are lots of other things to concentrate on fixing first). But in an environment where The Machine Simply Must Work, it’s still an unacceptable risk. And so the best practice for these environments is still assuming that any update that requires a reboot to complete should have that reboot performed as soon as possible.

Second, even setting the “Does the update do the right thing before the reboot?” problem aside, CBS [Component-Based Servicing] itself creates another problem in this scenario. Since many Windows updates wind up getting their processing delayed until the reboot, that reboot itself can take longer (since all those pending operations then get performed). And in an environment where The Machine Simply Must Work, this means that the consequences of an accidental/unplanned reboot can be even worse. So again, in this environment, it’s important to ensure that Windows Update never initiates a reboot on its own. And since Windows Update will sometimes initiate reboots on its own when it’s set to install updates automatically, this means that the best practice for these environments is still the practice described in the blog: Set Automatic Updates to not install updates automatically, and use your own code to install updates and reboot at the correct times and with the proper user notification.

To reiterate earlier caveats, when I talk about situations where The Machine Simply Must Work, I naturally presume you’re not talking about life-critical medical applications, because Windows is not for life-critical applications, as the esteemed attorneys who hand-crafted the Windows EULA from artisanal Unicode characters will happily point out.

With regard to that latter, @SwiftOnSecurity quips: “Microsoft would like to remind you that WINDOWS SHOULD NOT BE INSTALLED ON PACEMAKERS.” I’ll, um, keep that in mind.

Comments (2)




The maybe-fly list

I can’t say this is too awfully surprising:

It gets, if not better, at least a bit more complex:

That’s a whole lotta veggies, San Francisco to Amsterdam.

Comments (4)