"Going Sleeper"? I don't get it. If you were laid off, why wouldn't you say so? You're under no obligation to a retailer you no longer work for, least of all to use some kind of coded language.

Oh please...

The XP SP2 / Windows Firewall came out in 2005. Nobody was using Netware at that point; Novell never even put out a formally-for-XP iteration of their Netware client for Windows; firms still using Netware just installed the Windows 2000 version...which was just fine, because Netware had priced themselves out of their own market in comparison to Windows Server's everything-in-a-box solution, especially SBS server that was almost impossible for Netware to compete with.

Lotus Domino was undoubtedly better than IIS as a web server, but Microsoft had 90% or more of the browser market at the time. Firefox was only just starting to make inroads, and Netscape had a lower market share than Safari. ActiveX and the ASP.NET stack sucked from a security standpoint, but Lotus didn't have a viable alternative to provide. The one thing Lotus *did* have available was Lotus Notes, the Exchange competitor, and I've never once heard *anybody* describe the experience of Lotus Notes as anything other than one which made people yearn for Exchange.

Wordperfect was already in its niche in law firms but Word was the default word processor, Amiword was bundleware, Lotus 1-2-3 lost its shine in the 90's and had long since given its crown to Excel, Access and *maybe* Filemaker had ousted Paradox.

By 2005, Microsoft *did* own the business tech stack; let's not confuse 2005 with 1995.

See, here's the thing: I completely believe you. I absolutely believe that Google's data handling is quite possibly the best in the industry. I believe your statement that PCI Compliance (which is too onerous for a whole lot of main street businesses if you ask them) is too insecure for Google's IT folks. I believe that the models they implement are literally state of the art and that more companies should follow those practices.

And, in a depressing irony, it makes me trust Google even less.

Why? Because how how data is taken in the first place. Take an Android phone out of the box, attach it to Wi-Fi, go through the setup and decline every possible opt-out switch (which are all enabled by default, obviously) and watch your firewall logs. The amount of outbound traffic is ridiculous, and way more than a regular update server.
Try removing all Google apps from the phone. You can't without going through USB Debugging, and even then, there are the below-the-surface tentacles that require an absurd amount of research to understand what's actually being broken, and even then, you probably have to root it in order to remove it.
Set up a Google Account (which you *have* to do in order to install apps from the Play Store; which I'm inherently okay with in isolation), and see what happens: if there are calendar entries, contacts, tasks, or photos already on the phone, it uploads them without asking, unless you disable it, which you can only do *after* the sync started.
Attempt to write a tutorial to make an Android phone sync as little data as possible with a Google account - fitness data, location data, passwords, autocomplete data...any data that's possible to opt-out of in a newly-minted Google account, write a tutorial on disabling it. Even if I believed that those switches were respected and not simply "IF setting='disabled' THEN field.visible=0" indicators, see how many steps are required to opt out of Google data collection on Android. I doubt it could be done in less than 50 due to the maze of different privacy setting screens.
Even if someone were willing to go through that entire list, Android has a tendency to add new data collection functions in their mandatory OS updates (and invisible Play Services updates), and opt users in without so much as giving users front-and-center opt-in indicators when those services are added. One must instinctively know to go to the Google Account settings regularly and check for new things to opt out of, and just accept that whatever data has already been transmitted, has already been transmitted.

So yes, Google keeps all of this data safe, but they kinda have to because they collect so much more of it than most people truly desire to be collected, especially if it were to adversely affect the wrong person (Imagine the contacts list or synced photos of Lebron James, Martha Stewart, or Jay-Z getting leaked to the public).

Google may have the most secure datacenters in the world, but that data is being used *somewhere*, for *something*, with no transparency or oversight...and it's the not-knowing that makes the security of their datacenters irrelevant.

I trust my own, way-less-secure server to sync my data, because the /e/OS fork of Android allows me to sync to my own server, for which they provide an installation script and all of the source code. Now, I try to keep it pretty-secure (it's behind a Sophos firewall with lots of the UTM functions enabled, with unique, >12character passwords and fail2ban), but i'm sure the data would be even more secure against hackers in Google's datacenter than in my podunk little Poweredge server. Even so, I know exactly where the data lives and what it's being used for...and I'd rather a 0.01% chance of it getting hacked than a 100% chance of my data being fed into some AI algorithm at Google and divulged to anyone who writes a big enough check or subpoena.

The rewards are opt-in. The "hahaha" from the GP was presumably implying disbelief that the tracking and data brokering are opt-in.

Trust me, I've got no great love for government operations, either. To put an even greater point of concern with this plan, there's an extremely valid concern about the federal government publishing all of its source code; security-through-obscurity isn't an ironclad plan by any means, but telling the entire world what source code is being run the sort of thing that has the potential for consequences, since vulnerabilities would be published for adversaries to see. This is especially problematic if security patches end up moving at the speed of government.

That being said, the current situation is that my tax dollars going to Microsoft and Oracle and Cisco and 101 different middlemen all feeding at the government trough. Those companies get billions of dollars, and we get...healthcare.gov and ransomware payments buried in a line item for 'outside consultants' somewhere. If, instead, a Department of Software Development where my tax dollars are spent writing code that can then be used by everyone, public and private alike, such that even code written by those vendors and consultants is subject to being open sourced so there's always at least some return on investment.

The first variant doesn't even get me a used Cisco router. The second variant, theoretically, gives me the ability to compile code to make my own router if I want to.

It had something to do with it, sure, but I don't think it's the real issue. After all, most games within the past 5 years can be played at 45fps on 'medium' graphics settings, right? Sure, it's not the best experience, but it's a way to play the game with the aging GTX970 that's still in the machine until prices and availability come down. Moreover, eBay is flooded with GPUs now; even if they're coming out of a mining rig, it's easy to get a 2080 or 3060 for under $200 used, which is more than sufficient to play basically every game released on 'high', if not 'ultra'.

No, there are two real issues here: time and money. As TFS said, gamers have fewer hours to game, which means they're going to be more selective of where they play it. In addition, the money factor is a big one. $70 for a one-time, perpetual license is a number I'd accept, even $80 if it's truly one-time...but it's not anymore, because everything is always-online, with season passes, battle passes, multiple in-game currencies, and lootboxes. $70 is the cover fee to get into the casino, which is going to either require an obnoxious amount of grinding, or another pile of money to get to the 'fun part' of the game.
Even if a particular person finds themselves enjoying a game enough to be willing to spend some $250 on in-game items, that only lasts until the game is shut down because the game didn't make bazillions of dollars in the first quarter after release.

After a couple of times of having that money flushed down the drain as a result of a game getting shut down, that person is going to be selective about the next game that is going to be on the receiving end of both time and money. Eventually, without either a game to play nostalgically as can be seen in the vibrant retro gaming community, or some tangible element of the experience (one can take selfies with friends in a casino, that's awkward to do in a living room by one's self), the 'growth' element can't be increased through gameplay or story-driven elements, but instead ever-more-predatory monetization that demands more time and more money from players who have neither.

If a game isn't enjoyable to play at 'medium' graphics settings at 1080p, it's not enjoyable to play on 'ultra' in 8K or VR. Don't believe me, go back to the original Nintendo Wii - the least powerful console of its generation, and some of the most fun playing video games ever experienced.

Step 1: Download app before purchasing the product.
Step 2: Attempt to get to the "add new device" page without creating an account.

If Step 2 fails, the device is a poor investment. If it succeeds, it's probably safER to purchase (yes, I'm still salty that the new Kasa app mandates account creation when that wasn't a requirement at the time of purchase).

But really, smart home equipment is still in the experimental phase and should be treated as such. If the goal really is long-term usage, start with HomeAssistant or OpenHAB or Domoticz and go from there. The fact of the matter is that it's become pretty clear that Google and Amazon and even Apple have no *real* interest in making smart home integrations viable long term. The problem is that it unravels pretty quickly into a hobby of its own; most of this stuff is cheap enough that the majority of purchasers are more likely to just landfill stuff when the software artificially breaks, rather than try and salvage the equipment with Tasmota or ESPHome.

And it still exists, if you want to use it.

The problem is that a vanishingly small number of people do, because NNTP is a protocol reflective of its era, starting with the need for a dedicated client. Most clients avoid the worst elements of recent UI design (flat icons, ribbons, excessive whitespace, etc.), but go to the other extreme of all looking right at home in Windows 95. Usenet readers aren't the sort of things that are going to need frequent updates, but even so, more than half the GUI newsreaders haven't had an update prior to 2020, and those that have are mostly things like Claws Mail or Thunderbird, i.e. e-mail clients for which NNTP is basically an afterthought. Then, there's the fact that non-Windows, non-OSX devices like iPads and Chromebooks are severely limited; apps exist for both platforms but those solutions are more limiting than what is possible with Reddit (and was even more readily possible prior to the API lockdowns; no mobile NNTP client was as good as Apollo was).

Ignoring the issues with connectivity, Usenet lacks a whole lot of functions that improved the quality of life for most users. Upvotes and downvotes help highlight desirable content. Emojis and GIFs embedded images are generally popular and supported by Reddit. Text formatting options like bold and underlining aren't possible in Usenet posts, let alone colors or automatic lists. But the real problem is that Reddit's spam level and Usenet's spam level are two different things, and the difference would diverge considerably if Usenet ever experienced a renaissance. Similarly, there have been a whole lot of unfair bans, shadowbans, and moderation abuse...but the absence of *any* of these things in a community swings too far in the other direction, preventing any real ability to curtail genuine troublemakers.

So...yeah, Usenet existed, and continues to do so, but by time extensions are in place to reflect the internet as it is today rather than in 1986 when NNTP was first standardized, we'd end up with something almost identical to the very centralization that is seen as problematic, likely only different through the absence of actually-desirable features Reddit has.

Oh, and there's no such warning. When you get into the backseat of a robotaxi, it won't even start moving until everybody is wearing their seatbelts. I hardly see any passenger leaping into the front seat and grabbing the controls in a traffic incident.

Nah. I've rode a couple of robotaxis around the city, and while the ride is nice, it's clear we're nowhere near "get in a car and go to sleep." For one thing, the vehicles aren't even allowed on highways yet. And they require months of training on any particular urban area before they can perform reliably. I don't think there's been any training in rural or even suburban environments, which have different challenges. What you say may eventually be possible, but we're still a long way off.

Couple of reasons...

1.) Virtualbox is, as stated, a type-2 hypervisor. It's an application that sits on top of another OS, while VMWare ESXi is an OS itself. This puts it in a largely different category, for reasons that will be clearer later on...but it's at least worth noting that in all of the years since Oracle bought Sun *and* forked Red Hat Linux, nobody, either at Oracle or in the community at large, has created a Type-1 hypervisor out of Virtualbox.

2.) One of VMWare's main selling points is its management solutions. If you have a single host, log into the host's webUI and manage your VMs from a web browser (or the 'thick client' before it). If you have ten hosts, log into your VCenter server and manage your VMs and your hosts from the web browser. If you have a thousand hosts, VMWare has an API that allows management, either manual or automated, through a command line...in addition to the VCenter server. With Virtualbox...you have to use some sort of RDP/VNC connection into the host and make changes to your virtual machine that way. If you have 10 hosts...you have 10 VNC connections. While this is fine to replace VMWare Workstation, its not a very scalable management solution.

3.) Because Virtualbox is a type-2 hypervisor, it is dependent on its underlying OS for storage and networking. Want certain VMs on a VLAN? Create a port group with the VLAN tag and assign those VMs to the port group. Done. Want that done in Virtualbox? I hope your NIC enables VLAN tagging, or that you've got multiple NICs and a managed switch! How about storage, want to mount an iSCS LUN? Hope you know how to do that in your host OS *and* that you're cool with how the host OS handles iSCSI. Want to mount an NFS share? again, hope you're cool with how the host OS handles it!

3.) Vmotion. The ability to move VMs from one host to another, or one storage volume to another, without downtime, is impossible to replicate with Virtualbox. One would have to shut down the VM, mount a shared folder on the destination host, move the VM, and then start it up again on the new host. Oh, and since it's just a straight file copy on the host OS, it's competing with the other just-an-application (i.e. *the hypervisor*) for network bandwidth and disk I/O, so while VMWare will prioritize running VMs (unless you tell it not to), the host OS won't, and everything slows down until the move is complete.

4.) Want to back up your VMs? If you're on VMWare, there's Veeam, there's Altaro, there's BackupExec, there's Nakivo, there's Veritas NetBackup, and a dozen others. If you're on Virtualbox? ...I guess you can shadow copy the file system of the underlying OS and then do a bare metal backup of that, but a restore operation now means restoring that bare metal backup *elsewhere*, pulling out the VM in question, moving it over, and hoping that it's not corrupted, because I'm unaware of a backup utility that will error-check the VDI file.

5.) Because other vendors integrate with VMWare, and don't integrate with Virtualbox. For starters, you won't get per-VM load or utilization stats; what's a VM and what's a 7-zip operation are going to look basically-equivalent to most monitoring tools. The aforementioned backup applications don't support Type-2 hypervisors at all. If you need offboard storage, you *can* connect a Synology or a QNAP to the underlying OS, but God help you if you're trying to connect something from EMC or NetApp or even iXSystems. You *might* get it to work, but you're not getting support for that piece of equipment that cost more than a car.

6.) Because others exist. Again, Virtualbox is fine for having a little 'ant farm' on a desktop, or running an old application locally, or something to that effect...but other type-1 hypervisors exist, including Hyper-V, XCP, Proxmox, and Nutanix...while a type-1 iteration of Virtualbox doesn't.

Those are just a handful of reasons.

Well, I'm sure that Tan would like nothing more than for every VMWare customer who's ever existed to write him a $10,000,000 check, daily. He said the quiet part out loud, and likely too early on in the process, that the goal was to keep the top 600ish customers and squeeze everyone else out.

There's a subtle difference, I think: Tan is more likely to be unhappy about the fact that he doesn't get to pick which customers can easily dump their crap. In a world perfectly suited to Broadcom's wants, the VSphere Essentials customers and high-call-volume customers would be able to easily migrate elsewhere, because VMWare doesn't want them.

The bigger issue is that the companies VMWare *does* want to keep are the extremely large ones. Extremely large companies can afford skilled personnel. Those skilled personnel can spend their days moving at least *some* of the workload to other solutions, whether it be Hyper-V or Proxmox with its CLI tools, or Nutanix or XCP with conversion scripts or backup/recovery tools...the companies that are large enough to be the 'whales' Tan wants to keep, are large enough to be able to throw enough people at the problem to resolve the matter in a way that *doesn't* involve Tan getting those renewal checks written.

The Proxmox tool is unlikely to alter the calculus that much under these circumstances; it's more likely to help the smaller companies VMWare didn't want, and perhaps accelerate the timetable since the conversion will be simpler, but will it hurt VMWare? Only if VMWare suddenly wants the small and medium businesses back...and the odds of them being able to do that at this point are pretty remote, since it's become pretty clear that Broadcom's direction for VMWare is one that approximately zero customers see as acceptable.

Proxmox has been a name that has gotten a lot more traction over the past several months; between this migration tool and the rumored Veeam compatibility, it may accelerate its status as the go-to replacement for smaller VMWare deployments, but it was likely going to trend that way regardless.

Dave Plummer may have Tweeted about it recently...but his Youtube video on the topic is over three years old. His Youtube channel is particularly interesting, since it includes information regarding Task Manager and a number of other late-90's era things he coded that are still in use, so it's well worth the watch.

In terms of the disk format tool, there are a few things I'd add if it were up to me. First, I'd add a "low level format" option that nukes the partition table entirely and doesn't create a new one. I'm a happy purchaser of a license for LLF, but this really should be a built-in function, especially since it's technically possible to do with diskpart in the CLI. Second, I'd add a 'custom size' option that allows the user to specify the partition size. This is possible to do in the MMC snap-in (diskmgmt.msc), and of course diskpart allows it as well, but I think that the ability to make simple sized partitions would be a viable improvement.

That being said, I hope Microsoft doesn't mess with it, because Notepad is a harbinger of what is coming. Notepad was pretty darn close to the "do one thing and do it well" UNIX philosophy. It did text editing at a very predictable level, used next to no system resources to do so, and had basically four options: find/replace, word wrap, a font format dialog that changed how Notepad displayed whatever it opened but didn't format the text at all, and the F5 option that automatically entered the date/time...and that's it. Now, the devs decided it needed to be 'modern', so it got tabs and 'search bing for' and autocorrect and spell check...who thought it was a good idea to have autocorrect in a text editor?!?

If this is what Microsoft is doing to legacy code that has stood the test of time, my hope is that they don't mess with it, because there is no possible way that the current folks at Microsoft will be able to improve upon Dave's code. Knowing them, they'll have Clippy Cortana Copilot integrated and suggest partition layouts based on what's trending on the Bing Start Page.

But really...R.I.P. PartitionMagic.

Unfortunately, this is getting harder and harder to implement. Based on your Slashdot ID (roughly half of mine, and I signed up in 2008), I'll wager that your first few years of computer use involved using computers that didn't have internet connections. We had games on cassettes (if you're older than me), floppy disks (if you're my age), and/or CD-ROMs (if you're younger than me), which made computers useful in a standalone capacity.

I'm hard pressed to think of a game or software title from the past few years, intended for children, which retained that paradigm. I'm sure there are a handful of iPhone games that may be one-off purchases and can be played without an internet connection, but even the later iterations of Angry Birds mandate an internet connection, even on the paid iterations and where gameplay does not depend on a server component. My nephew likes Roblox, my first-cousin-once-removed likes Fortnite, and my niece likes this fashion choice game where, shockingly, most of the outfits are in-app purchases.

Even if you're the kind of parent who's going to ensure your child uses Edbuntu, their friends won't. Even if you're the kind of parent who plans on giving their child a laptop or desktop to use for their digital learning, you can't load Edbuntu on a hand-me-down iPad.

The depressing reality we find ourselves in is one where preventing children from being online unsupervised is basically to keep them from using computing devices. An analog childhood is a good thing for most children, but it requires incredible dedication from parents to implement, doubly so in finding a community of similarly-minded parents, lest your efforts be undone as children meet friends who enable them to have unsupervised internet access.

Because that worked so well for Iran.