Stories
Slash Boxes
Comments

SoylentNews is people

posted by janrinok on Saturday January 11 2020, @04:32PM   Printer-friendly
from the only-the-date dept.

Windows 7 and Server 2008 end of support: What will change on 14 January?:

It is remarkable that Windows 7 is reaching end of support on January 14 2020 while maintaining something approaching 27 per cent market share among Windows users, according to Statcounter.

This is down from 35 per cent in December 2018 but still substantial. Windows has a share among desktop users of around 77 per cent, so that is around 20 per cent of active desktop PCs.

"End of support" means no technical support, software updates or security fixes from Microsoft. Of these, the significant piece is the security fixes. Without regular patches, flaws that are discovered in the operating system will put users at greater risk from things like ransomware attacks, perhaps triggered by an email attachment or malicious web link.

Windows Server 2008 and 2008 R2 also go out of support on the same day. Although it is less likely that users will be browsing the web or clicking attachments on Server 2008, it is still risky if these servers are exposed to the internet – as appears to be the case with Travelex, currently suffering a ransomware attack – or if they are used for remote desktop services.

Another curious feature of this "end of support" is that Microsoft will still be providing security updates for both operating systems, for three further years. So the real end of support date is in 2023. That said, you can only get these "extended security updates", or ESU, in certain ways:

  • Windows Virtual Desktop (WVD) users get free ESU until January, 2023
  • You can purchase Windows 7 ESU by subscription from Microsoft Cloud Solution Providers, which means most IT support companies signed up as authorised Microsoft suppliers.
  • Windows 7 ESU is free for a year to customers who subscribe to Windows E5 or Microsoft 365 E5. Details are here
  • Only Windows 7 Professional and Enterprise are covered by ESU.
  • Windows 7 embedded can be supported through an "Ecosystem Partner Offering" support contract.
  • The scenario for Windows Server 2008 ESU is similar to that for Windows 7.

[...] There is a degree of artificiality about this key "end of support" date and ways to keep old stuff patched, but the security risks are real.


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 2) by RS3 on Sunday January 12 2020, @05:06AM (2 children)

    by RS3 (6367) on Sunday January 12 2020, @05:06AM (#942415)

    Yes, awesome post. I replied before reading yours.

    I think I've seen more power supply (the metal box) caps go bad than motherboard. But yes, many otherwise good PCs were trashed, and I'm pretty sure it was Dell who ended up replacing tons of PCs due to the caps, but it plagued all manufacturers. I think Asus and likely a few others were more careful to use good caps, but I'm sure they all got hit with defective caps with faked good-brand labels.

    I still need to buy / build an ESR (capacitor) tester. I've found and replaced many bad caps that had no outward signs. Just loaded the PS and looked at the voltage with a 'scope.

    And it's not just computers- I've fixed tons of TVs and monitors with bad caps. And all kinds of wall-warts, line-lumps, etc.

    To be fair, even good caps used in switching supplies have a fairly limited lifetime spec. Personally I'm not thrilled with it. IIRC there was a story here on SN about someone who figured out how to make switching supply caps last much longer by using non-electrolytic caps along with electrolytics.

    Starting Score:    1  point
    Karma-Bonus Modifier   +1  

    Total Score:   2  
  • (Score: 2) by Grishnakh on Sunday January 12 2020, @10:43PM

    by Grishnakh (2831) on Sunday January 12 2020, @10:43PM (#942570)

    >To be fair, even good caps used in switching supplies have a fairly limited lifetime spec. Personally I'm not thrilled with it. IIRC there was a story here on SN about someone who figured out how to make switching supply caps last much longer by using non-electrolytic caps along with electrolytics.

    Electrolytics are really pretty awful capacitors in many ways; the problem is that we don't have anything better that doesn't cost a small fortune (which is tantalum caps, which also are made of conflict materials). If you read the specs for them, they have a lifetime rating in hours, usually only a few thousand. If you think about it, that's not much time for something that gets used a lot, or worse is on 24/7. Of course, that rating is generally worst-case, so if you run it at non-worst-case conditions then its lifetime goes up, but still: ceramic capacitors don't have this limitation.

    High-end equipment will frequently be made with tantalum caps, which are a little smaller, and last longer, though they're not nearly as tolerant of overvoltage as electrolytics (they tend to explode if pushed past their ratings). And MLCC (multi-layer ceramic) is always getting better, so that's been taking over for electrolytics in some applications. Traditionally, ceramics had very low capacitance ratings, so were used for decoupling/noise, whereas electrolytics had far higher capacitance values, so could be used for power supplies, but MLCCs have been pushing values much higher. So, for instance, in the "old days" ceramics might have values in the pico and nanofarads only, and you'd use electrolytics for things from 1uF and up. But now, you'd only use ceramics for 0.1uF (100nF), but they also come in 1uF and 10uF commonly, with the latter being firmly in the electrolytic range a couple decades ago. So there's really less and less need for electrolytics these days; only for the really huge values (like over 1000uF).

    As for your story, I don't think I ever saw it, and IINAPSE (I'm not a power supply engineer), but if I had to guess on a quick way to improve PS performance and reliability, it'd be to use MLCC caps in parallel with large electrolytics, with a couple different values (perhaps 0.1uF and 10uF) of the MLCC, so that fast transients are passed by the ceramics and not the electrolytics. But again while I might sound like I know what I'm talking about here, I really don't; I'm an EE but I've been working mostly in software most of my career.

  • (Score: 2) by Grishnakh on Sunday January 12 2020, @10:47PM

    by Grishnakh (2831) on Sunday January 12 2020, @10:47PM (#942572)

    >And all kinds of wall-warts, line-lumps, etc.

    Personally, I wouldn't even bother fixing these any more, unless it's already a switching PS and you can get the replacement caps cheaply and easily. If it's something older, I'd rather replace it with a newer switching adapter. Those old wall-warts with the iron-core transformers had absolutely terrible efficiency, and wasted a lot of power even when not in use, just being plugged in. They were even called "energy vampires" because they accounted for a surprisingly significant portion of the load on the grid, for such small devices. The new ones are much, much better, and use almost negligible power at no-load conditions.