If your PC doesn’t run Streaming Single Instructions Multiple Data (SIMD) Extensions 2, you apparently won’t be getting any more Windows 7 patches. At least, that’s what I infer from some clandestine Knowledge Base documentation changes made in the past few days.
Even though Microsoft says it’s supporting Win7 until January 14, 2020, if you have an older machine – including any Pentium III – you’ve been blocked, and there’s nothing you can do about it.
While support has to end somewhere – processors without SSE2 are really, really old – it’s quite unfair to say you support Windows 7 until 2020, and then cut it off early for a number of customers. Consumer protection agencies should have something to say about this, right?
Although it doesn’t affect me directly, it’s crap like this and Windows 10 that’s moving me over to Linux more and more!
My main machine is Windows 7 and multi boot with 3 Linux distro’s, Windows 7 will remain on it for a few years yet but it’s the last version of Windows I will run.
Things are becoming more anti consumer every week, Windows will probably be subscription only before long.
I think you’re right, but I would prefer it to the shit they’re doing now, like ads in the start menu and putting crap on the system that I can’t remove. If they cut all that out, I’ll happily pay a subscription. I know ‘subscription’ is a dirty word to most, but it really doesn’t bother me. Not that I want to pay $5 a month for every app I use, but I’ll tolerate it an the OS. I don’t expect to get everything for free.
And before somebody asks, ‘well, what about Linux?’ No, just… no. Desktop Linux was irrelevant 20 years ago when people started talking about ‘the year of Linux on the desktop’, is irrelevant now, and will be irrelevant 20 years from now. Because, for better or worse, the community around it seems to prefer it being the clusterf–k that it is, ensuring that it’s never going to become popular with the mainstream.
That being said, it might sneak in the back door with something like Chrome OS running native Linux apps, which is fine with me. I’d love a good Windows alternative that isn’t Mac OS
Edited 2018-06-23 15:04 UTC
Linux can be a pain that’s true. I don’t find it as bad as you make out though, it’s definitely more useable than Windows 10 for me, at least the XFCE DE is.
Unfortunately there’s nowhere else to go, unless Haiku performs a miracle and releases a stable version 1.0 soon that gains traction and users/developers!
Haiku needs to improve the driver situation first, or more specifically, the Wi-Fi drivers situation. Only then it’ll become more relevant IMHO and only then I can use it for more things than to play around with in a VM (I’d actually install it on this new laptop to dual-boot with Linux!).
Just use a supported WLAN card as there likely is one… most if not all laptops support swapping cards (though some like lenovo’s will require you to choose one on it’s whitelist).
That’s not a solution IMHO. If drivers exist for all platforms, including Linux and BSD, then Haiku should just support those drivers. I don’t want to open up my laptop and swap cards (plus I don’t even have money for a different card – this new laptop was already slightly over my budget but I really needed it).
“I don’t want to open up my laptop and swap cards”
But you expect the Haiku developers to do all the work for you… a wifi card is somewhere between $5-20 on ebay depending on the model and will solve your problems now rather than waiting for the developers to solve it while your laptop becomes obsolete. And honestly you aren’t likely to gain much sympathy with technically minded people with an attitude like the one are you are displaying.
While yes the Haiku developers are porting new drivers from current FreeBSD as we speak chances are the ones that are already working are going to be more stable for sometime to come.
To be on the safe side and prevent wasting $ you could ask the Haiku developers which cards they have working reliably today and see if one of those will work with your laptop.
Edited 2018-06-24 18:52 UTC
“And honestly you aren’t likely to gain much sympathy with technically minded people with an attitude like the one are you are displaying. ”
I already have gained sympathy within the Linux community, so unless you don’t count the Linux community as “technology-minded”, you’re dead wrong.
Anyway: I don’t see what the big deal is by waiting for drivers. On Linux, if something’s not working, we also wait for drivers to come.
Funny part was that one with money – you can get wifi USB dongle for 1-2$.
Funny thing is that Wi-Fi USB dongles don’t work with Haiku yet: https://discuss.haiku-os.org/t/haiku-beginner-sound-and-wifi-program…
Please do some research before you post witty responses.
I remember a time when RTC modem used the at-hayes command set to work through serial port, no needs of “driver” using a standardized command set. Same with dot matrix printers. Why standardized command sets are the thing anymore ?
Kochise,
I suspect “AT” instructions were a pseudo standard, and devices implementing them were clones. This kind of compatibility was very useful to be sure and fortunately was acceptable practice back then. Yet in today’s legal climate, implementing a device with API compatibility could very likely land a lawsuit due to the precedent set by the google/oracle case, which means we could get in trouble for reimplementing a compatible API unless we obtain a license first.
Those are just APIs, a set of “semi-structured” datas. This deosn’t reveal the underlying algorithms… Do’h…
Though it’s a trend which clearly precedes API copyrightability… Todays devices do quite a bit more while being much less expensive …probably thanks to a lot of functionality being moved to software, like winmodems, so it’s understandable that manufacturers want to block clones…
Subscription doesn’t have to be a bad thing. My Linux distro of choice, Solus, is (like most distros) free of charge, but I choose to donate $10 every month through Patreon because I want to support them. So while it’s technically not a subscription, it does kind of feel like that because I choose to “pay” for the OS every month
Edited 2018-06-23 15:26 UTC
Donating to a project and renting an OS is two different things. You can be sure Microsoft will lock things down even more and make sure you can’t do anything they disapprove of while collecting their rent at the same time!
I’m not sure you understood my post as I clearly said that they are two different things. So no need to repeat me. I just said that to me it *kind of FEELS* like a subscription because I donate every month
We all understood it. But you prefaced it with ‘subscription doesn’t have to be a bad thing’ and then proceeded to describe something that is not in any way a subscription.
To be honest, I wouldn’t even pay 10€ per year for it.
Sorry, but the OS is just an extended bootloader for the application I want to use.
I really wish Windows would go back to being a simple resource manager, boot loader, and application manager. There’s really no reason for it to need GB of disk space for that.
Edited 2018-06-23 18:55 UTC
I hear a lot of complaint about Windows.
I have a second hand HP Z620 workstation as my Home Office machine. I can plenty of VMS on Windows, and a browser and Visual Studio.
Everything works fine on Win 10.
The OS has a lot of features and drivers, they take up disk space. There are alternatives if you wish to use them.
Edited 2018-06-23 23:53 UTC
Elementary OS has become stable enough that I use it as a daily driver now, moving Slackware to a hobby/fun OS. I no longer run Windows 10 on this workstation at all, only on my HTPC/gaming rig. Wine is good enough at this point to run any Windows app I’ve thrown at it, and most of my workflow is platform-agnostic anyway (basically I just need a web browser and a terminal).
I get that for some folks, Linux on the desktop will never be “good enough” and I can respect that, but I think it has definitely reached the point where the OS moves to the background and the user can just get things done without fiddling around.
I don’t think an OS is something a person should have to subscribe to. I despise the idea of being strong-armed into monthly payments or be spammed with advertisements and garbage. I understand times change and not all business models are sustainable, but it seems the go-to solution is to spam the hell out of the customer with advertisements until they become submissive. If anything is hostile towards the user, it’s that.
Sure, I get it. But even as you eluded to, I don’t think a mammoth release every 3-5 years is sustainable anymore, so unless you expect developers to work for free, probably the only other alternative (assuming you’re not making money on the hardware side like Apple does) is a release every year or so, which you’d have to pay for anyway in order to stay current. So, pick your poison?
Honestly, that’s justification enough right there. Change can be hard and some of us have to kick & scream a little to get through it. I feel pretty sour about a lot of stuff happening these days but at the end of the day I know these are just the times I’m living in.
So you support piracy, yes? Because I hate to break the news to ya but systems using CPUs that are THAT old running Win 7? 99.995% are running Win 7 Ultimate, AKA “Windows Pirate Edition”.
I’ve worked in PC shops and Home Theater installations for nearly 25 years and I have NEVAR seen a system THAT old running a legit copy of Windows, not once. Pirates will take an old POS, slap their Windows Pirate Edition on it, and turn it for a little cash, why should MSFT spend a pile of money on QA on systems that no OEM ever sold that I’m sure their own data shows is nearly all pirate boxes?
Even Linux has limits on how damned old a CPU you can run with a modern kernel, expecting MSFT to support ancient junk that is running stolen copies of their product? Kinda being a bit hypocritical, don’t ya think?
BTW if anybody is wondering since you can run Win 10 in limited mode without a key? The pirates have already started moving over to Windows 10, in fact last time I looked at my local CL there was multiple ancient boxes running windows 10, the oldest was an Athlon X64 X2. Now does anybody here REALLY believe that that $50 PC tower on CL is running a $100 copy of Win 10 Home legit? Yeah…no, its either in limited mode or Windows Pirate Edition from TPB.
Yep have to completely agree.
There is a whole Youtube Channel dedicated to it called
“RandomGamingHD” where he tests out a lot of juck systems or builds a super budget system using second hand parts.
I don’t know where are they still finding this trash Pee-Cees and complaining about vendors not wanting to support their 20 year old garbage. I checked second hand PCs market just now and you can even buy old Core i3 HPs and Dells for less than $75. These ancient PCs are perfectly capable or running even windows 10, however I doubt anyone is that stupid to spend $200 on Windows 10 home and then running it on something that people don’t want even for free.
Crap like a 9 year old OS not keeping support for 16 year old hardware.
Get real FFS.
I think the issue is that Windows 7 had minimum requirements and claimed to be supported until 2020. They changed those requirements before the end of the support period. It is a bit like breaking a social contract.
This is not Windows 10.
It’s a non-issue then. The OS that was purchased 9 years ago will continue to run on the 17 year-old hardware as it did yesterday. It’s just not getting updates. If one can live with crappy old hardware, surely one can live without updates.
It is a bit of a non-issue IMHO.
The applications that can run on these processors are becoming increasingly rare even if they can run at all.
What is the point of a OS security update if almost all of the applications are insecure or the performance soo poor they aren’t worth using.
Eh, it happens to me on linux too, just more randomly and not on purpose. I have some obscure hardware apparently. I had to switch off Fedora because a kernel update broke its ability to boot..
I kept an old kernel around for a while, hoping that it would get fixed, but eventually I had to just migrate to ubuntu which apparently doesn’t have what ever change that was to what ever driver that broke me. I don’t really have time to do a diff on kernel modules, or source code between the two to see what it is. I guess that’s my problem as I haven’t paid either RH or Canonical.
It probably didn’t happen by purpose at Microsoft either.
Why do people automatically assume the worst?
https://en.wikipedia.org/wiki/Hanlon%27s_razor
I can tell you as someone that works in very large organisations. Most things like this are cockups and people trying to cover their arse.
What I suspect happened is that someone screwed up, it wasn’t tested properly and management are basically saying “This isn’t a massive priority”.
Possible that it was done without CPU consideration. I can see a manager deciding to not rewrite code for older cpus like that. Pentium 3 is old.
These are pentium 3 machines, they are over 16 years old now. Most modern applications won’t even run on these machines.
e.g. https://support.mozilla.org/en-US/kb/your-hardware-no-longer-support…
Will this actually affect a large consumer base? No. So who would be getting protected? A very small amount of people with what is now ancient hardware.
Edited 2018-06-23 17:34 UTC
Had a look, SSE2 was supported starting on pentium 4 which was released in nov 2000. So for Pentium 3 you are talking 18 years ago which would be hard pressed to run and video content to start with now days or even be able to run code on modern sites without being super slow.
The last pentium 3 chip about 2001, with a final 1.4 ghz processor that was actually faster clock for clock than the P4.
I am sure you can find the model / die (?spelling) names on wikipedia somewhere.
That period was when I first started working and upgrading PCs. So I know it well.
These machines are so defunct now, while I am sure there are a few of use that respect them for how good they were back in their time. They aren’t used by anyone for any serious work.
Edited 2018-06-23 23:32 UTC
So basically, Windows 7 is dropping support for a generation of chips that was already out of manufacturing 8 years before that version of Windows was released.
Yeah and apparently some people think it’s outrageous.
All Pentium 3 CPUs had higher IPC than any P4; and that last P3 1.4 GHz Tualatin was faster also in absolute numbers tham quite a few higher clocked P4s…
For me, Windows 7 is the most recent version of Windows I can tolerate. I feel somewhat in a minority when I say that Windows 8 and 10 just feel like they’re ‘getting in my way’. Most of the other power-users I interact with love it to bits.
That said, I don’t think many of us run Windows 7 on hardware that old. It wouldn’t make much sense.
As soon as Windows 7 security updates die I think I’ll be switching Windows off for good.
Perhaps they should tie in ongoing support (at least security updates, and ideally feature updates too) with published minimum specs for that Windows version (and with added features for later builds or later service packs – maybe it’s OK such published specs creep up with newer builds also)
..but getting back to the point: for “old” CPUs, and I’d say >8 yo, maybe MS could take an honest tack of suggesting, if they can’t recommend, that owners of these older computers use those only for offline use.
Or if they still wish to use online to move to various free software distributions that better support or just support: secure builds of their OSes. I’m thinking Linux, NetBSD and the like.
… wishful thinking I’m sure. Ethics in large corporations huh. Oxymoron..
🙁
People here seem to have a strange idea about what forms the major base of PC users. Take a look at something like Steam’s stat page – you’ll see that the vast majority of users are using CPUs that are six to ten years old. While dropping twenty year-old hardware is fine, ten years on the hardware should be a minimum, with fifteen being better. Given those terms, dropping P3 support shouldn’t be a real issue. As other posts have mentioned, that chip is at least 17 years old now.
Microsoft -hates- that. That’s why they’re pushing everyone to Windows 10, and will then push an update that essentially cryptolockers your PC via bitlocker if you don’t have a valid subscription key.
How did you infer that? Steam hardware survey doesn’t list models/names of CPUs, only number of cores ( https://store.steampowered.com/hwsurvey/cpus/ most people are on 4 cores, which 6-10 years ago was still relatively rare IIRC) or frequencies ( https://store.steampowered.com/hwsurvey/processormfg/ which seem recent). At least models of videocards are listed ( https://store.steampowered.com/hwsurvey/videocard/ ), and they largely seem quite new…
PS. Though Steam users partly self-select themselves to have hardware on the more powerful side of the distribution curve, so the average PC in Steam stats is probably somewhat more powerful than real average / I agree with your general sentiment, many people use somewhat old PCs…
Edited 2018-06-27 00:27 UTC
Reading through the linked articles, there a few things which were not surprising:
-Patches are rolled-out which break systems meeting the official minimum requirements for Windows 7.
-Customers complaints and a “We are working on the issue” message distributed.
-The issue is swept under the rug and broken systems (no-boot) are left stranded.
While PIII systems are ancient, if they met the minimum requirements on installation of the OS, they should not be dropped from support before the EOL for this OS has been reached.
Maybe Microsoft could not test the patches before release, and cannot develop a fix, because there are no PIII systems left anywhere in the organization?
If all the PIII systems were recycled, would a company dedicated to customer support have realized this and gone on the secondary market to purchase some systems at the minimum requirements for the OS for testing of patches before release?
Consumer protection agencies should have something to say about this, right?
why?
if anything, they should say something in the line of: “get newer f–king hardware” or even better:”thank you MS, for supporting chips that were out of production for 17 years (as of today)”.
and I know it’s somehow still popular among pimple-faced “l33t h4xx0r5” to bash mainstream windows, but FFS, grow up.
If you still use ancient hardware for professional tasks than you have much bigger problems than end of Win7 support; andd if you’re using ancient hardware for hobby purposes, then why are you even whining?
edit:typos
Edited 2018-06-24 07:30 UTC
Still use old hardware with the suitable operating system that get the job done : VIA C7 = 2000/XP, Intel Atom 330 and AMD E350 = 7. Sadly it’s already too much for the Atom 300 because limited to 2GB of memory.
So those who have “bigger problems” can deal with one more worry? So it doesn’t matter how much we break for them because they’re useless/dead to us anyways? I don’t mean to be snarky, but if anything, the focus should be on helping the weak, not the strong.
Newer machines don’t need the added effort, only the older ones. “You’re only as strong as your weakest link.” Heck, optimizing for older hardware often improves speed on newer hardware.
Granted, there are many other hidden problems, so it’s not so obvious. I’m just saying, maybe don’t always focus so much on optimizing for hardware that isn’t even manufactured yet? New and shiny always attracts attention, but old (which was once also new and shiny) is too often ignored whenever something new comes out. People who have lived through dozens of such releases should still have some sympathy for “old” (e.g. 2011 or whenever). But they don’t.
We’re constantly filling landfills with working hardware. At some point we’ve jumped the shark. Good enough is good enough. (Or do you think everything should be “AVX only”?? 2011 was like seven years ago, OMG!! That’s two college degrees!! Gimme a break.)
@rugxulo:
So those who have “bigger problems” can deal with one more worry? So it doesn’t matter how much we break for them because they’re useless/dead to us anyways? I don’t mean to be snarky, but if anything, the focus should be on helping the weak, not the strong.
if one still uses abacus for heavy calculations, than the lack of replacement beads is not their biggest problem.
Newer machines don’t need the added effort, only the older ones. “You’re only as strong as your weakest link.” Heck, optimizing for older hardware often improves speed on newer hardware.
if you have enough money to spend for optimization. IF…
I’m just saying, maybe don’t always focus so much on optimizing for hardware that isn’t even manufactured yet?
optimizing for chips that were outdated 17 years ago makes even less sense.
New and shiny always attracts attention, but old (which was once also new and shiny) is too often ignored whenever something new comes out. People who have lived through dozens of such releases should still have [i]some sympathy for “old” (e.g. 2011 or whenever). But they don’t.[/i]
I have sympathy for older hardware, that’s why I still keep my “XP-Box” in excellent state – but that’s my hobby PC for older “win98 to XP era” games. at work I need the latest and greatest because older costs shit ton of time and therefore money. I could technically setup an old Win95 box, install “ye olde” arcview 3.0, use one day for setting up my data and then another day for calculating the data; OR I could do all that in couple of hours and save me at least a day and a half. Nostalgia is fun, but it has no place in professional environment.
2011 was like seven years ago, OMG!! That’s two college degrees!! Gimme a break.)
we’re talking about processor that was outdated 17 years ago; not 7 years ago. gimme a break, indeed.
Edited 2018-06-25 19:10 UTC
This is NOT an anti-user or hostile act. Ancient hardware can’t be supported indefinitely. You have to draw the line somewhere, and drawing it at CPU’s that were EOL’ed 17 or 18 years ago in software that’s a couple generations old is certainly well within reason. What other products can you name where you get nearly 20 years of support *after* the product was discontinued?
Hardware and software both have a shelf life. Nobody should be getting their tighty whitey’s in a twist because support is being dropped long after it should’ve been. The appropriate response is the opposite. You should be thankful for getting all those years of continued support, and move on. The only thing you accomplish by throwing a fit over this is making yourself look like an whiny a**hole.
This isn’t really up for debate. Even with the best arguments, it’s unlikely that anyone at MS (or anywhere else) cares what we think. So it’s pointless arguing here. But ….
So many things break these days, but nobody cares. Is it a virtue to abandon working hardware/software? “Newer is better! It’s faster, you’ll get more done!!” Maybe, maybe not. My point is that if you break it, it’s your fault. Better to not be the weak link that breaks the chain, if at all possible. Assuming that someone else will clean up your mess or fix your mistake is reckless. Make things better, not worse. Don’t just excuse and write off every little breakage. Make sure you aren’t the main one causing the problem for petty reasons. Let’s not blindly increase other peoples’ burdens, okay?
A P4/Athlon won’t even match a base model mobile Celeron in performance – and uses 10-20x as much power.
Now, now, you’re exageratting – I keep an old Athlon XP around, it uses max 65 W of power; which was typical for Athlons; from https://en.wikipedia.org/wiki/List_of_Intel_Celeron_microprocessors#… ~max 10x (and for quite a few Celerons ~5x, or even only ~2x) the power of current mobile Celerons.
It’s not about `faster is better` or `the new shiney`. There has to be a line drawn somewhere because businesses can’t ignore diminished returns. It’s not just unreasonable to expect support for this old stuff well beyond its shelf life, it’s unsustainable from a business point of view. That’s not the only reason to kill off old hardware. As another user pointed out, the really old stuff has a tendency to be far more power-hungry. While keeping a single machine running might not seem like a big deal, do the math if 100,000 people do it. Or 1,000,00, or 100,000,000+. It gets really bad really fast.
Why bother switching from incandescent bulbs to cfl’s or better yet to led’s? Why stop driving an old car that still runs fine but gets 10 miles to the gallon going downhill and blows toxic emissions like there’s no tomorrow?
Yes, some people are die-hard’s who share your point of view. I used to be one, and extremely stubborn about it. But in all honesty I can’t say having that attitude ever made anything better. I was more of a problem than a solution.
I’ll have to upgrade my old desktop! no longer will i have the time to harvest, kiln-dry, grind, perc and enjoy my coffee as my machine boots up! This is an outrage!
I’m not a Windows user, so care little about Win7 or it’s support model. I have a dual celeron 500 (Abit BP6) with 3T of raid 5 for offline backup, running OpenVault. It turns on by WOL and script shuts down, I’m quite proud of it. WTF would I buy new hardware to do something basic like this? I’m no environmentalist but I’m not buying more crap when I can use some thought to reuse what I have. There are usage cases for old hardware.
If you’re running these processors still, that’s the least of your worries. We are talking at 2004 release date for the newer versions, and these suck up so much power and the fans would have given out by now a couple of times.
I will admit though, it is crappy to change the terms later and the USA only has Corporate Protection Agencies now. It’s bizarro government now with 45.
And they were already left behind many many years ago. I had one as late as 2010, and while the windows kernel supported it and could boot, most applications required SSE2.
Hm, I was still using quite actvely a non-SSE2 CPU as late as 2015, and didn’t have any issues; browsers, video, audio playback software, compression software, TrueCrypt worked fine…
Fans aren’t that failure-prone …I have 2 from early 2000s, working fine / only one fan ever kinda broke for me (it still worked, just loud) …what failed was motherboard capacitors, they culd be a problem (and the replacement mobo I got from my buddy didn’t have mounting holes for big Zalman fan, so… 2 fans)
Windows 7 is now 8 years old and is still supported. That’s way longer than any warranty you can get on most consumer products. I don’t get this whining — is Microsoft somehow obligated to support it’s products indefinitely? Did it say anywhere on the retail box on your Windows 7 DVD “10 years of free unconditional support included”?
So were Windows 2000 and XP that I still use. Because they appears to performs just as well as when they were bought, and do 99% of the things that Windows 7 or Windows 10 do : UI, file managing, network, driver, 3D, application, web browsing, 3D rendering, video display, …
Sometimes I wonder what the next Windows versions have provided specifically, instead to remove features and add telemetry. What can we do now that we couldn’t before. Beside doing them “faster” and with a better graphic resolution.
Well, for me the only motivation to move on from Windows XP to Windows 7 was proper support for large amounts of RAM (>4GB). PAE on Windows Server 2003 borke some of my HW drivers and Windows XP 64-bit edition wasn’t really any better in that regard.
Then to move from Windows 7 to Windows 10 the only motivation was high-DPI support, to use 4k monitor.
If not for a lack of >4GB RAM and high DPI support I would still be running Windows XP very happily. Might have to use older versions of some programs, sure, but I don’t see a problem with that, as Windows XP was way better than Windows 10 after customising it.