It's time to seriously think about this. Windows XP is almost 15 years old, and is the last operating system to run pretty much any 9x software.

Sure it uses the NT kernel, but it's so old that it doesn't really matter (like 2000)

  • 1
    Windows XP was "was released to manufacturing on August 24, 2001, and generally released for retail sale on October 25, 2001." That makes it over 15, and almost 16 years old. – Michael Hampton Feb 22 '17 at 3:51
  • "Sure it uses the NT kernel, but it's so old that it doesn't really matter (like 2000)." That statement is not correct. There are very few changes at the kernel level between Windows XP and Windows 10. Obviously there are some, but they can be listed in a reasonably-sized list, and they are not fundamental architectural differences. Windows NT's kernel is still Windows NT's kernel, even nearly 25 years later. The vast majority of what you would read in Helen Custer's original Inside Windows NT (covering the original version, NT 3.1) is still going to apply today. – Cody Gray Feb 28 '17 at 8:25

It's still got 5+% of the desktop market share. Regardless of how old it is, I think that pretty well disqualifies it.

  • 4
    Agreeed, XP is still current technology. – Chenmunka Mod Feb 16 '17 at 10:56
  • 4
    While I agree that XP is current (or somewhat current), I don't think the percentage of market share should affect the choice at all. For example, can you tell me what the percentage of market share is for Amiga OS? While I'm sure it's in the subatomic fraction of 1 %, it still has "market share". So, what percentage points do we say is relevant to this discussion? Which is why I think market share shouldn't be a factor. – cbmeeks Feb 16 '17 at 13:25
  • 3
    I agree that market share shouldn't be a direct primary criterion. However, I'm not sure it should be entirely ignored. I would suggest, as a first cut at trying to define proper criteria, that it be considered along with other criteria such as manufacturer support and third-party support. For example, if there is no support from the manufacturer, there is no new software being published, and the "active installed base"/market share is below (arbitrarily) 3% and has been for (arbitrarily) five years, then it's "retrocomputable". – Jeff Zeitlin Feb 16 '17 at 13:46
  • 1
    (Obviously, those numbers are open to discussion, even aside from whether the basic criteria are appropriate.) – Jeff Zeitlin Feb 16 '17 at 13:46
  • I think that's what we need to have defined. But what do you mean by no new software being published? Do you mean from the manufacturer? Mean, no new XP compatible software from Microsoft? Because many new Windows apps will still run under XP. – cbmeeks Feb 16 '17 at 14:51
  • 1
    Instead of criteria such as x% market share, which violates the zero-one-infinity rule, or "no new software being published," why not leave it up to each individual to decide whether they think a Q&A site named "Retrocomputing StackExchange" will give them the answer they're looking for? – snips-n-snails Feb 16 '17 at 20:10
  • 2
    There are plenty of zero-one-infinity violations in, for example, law. One is that patents last 20 years; any patent-expired platform is more likely to be considered "retro". – Damian Yerrick Feb 17 '17 at 5:34
  • @Damian Whether a computer's patents have expired would certainly be a zero-one-infinity-conformant criterion. – snips-n-snails Feb 17 '17 at 18:53
  • 1
    @cbmeeks - by "no new software being published", I mean either that nobody is still publishing new software (as opposed to re-released) explicitly advertised as being XP-compatible (whether or not it actually runs on XP), or that any new software is known to be not compatible with XP. – Jeff Zeitlin Feb 21 '17 at 14:34
  • I wonder, though, about the actual purpose of the question. What if it is considered retrocomputing? What changes for the asker? If I had the privilege, the question would have attracted an "unclear what you're asking" close-vote. – peter ferrie Feb 21 '17 at 21:00
  • XP market share is more complicated than that. From the same linked Wikipedia page: "Web analysis shows significant variation in different parts of the world. For example, in North America usage of Windows XP has dropped to 2.06%, but in Africa it is still at 11.02%[62] and Asia at 6.63%[63] (even higher in China, but down to down to 18.21%, tied with Win10 at second[64]) while going down, and has been overtaken by Windows 10." – Michael Hampton Feb 22 '17 at 3:48
  • Maybe not so much "market share" as a raw number, but how much it's still deployed in business and industry as a main tool, not considering embedded devices or control systems. I've seen several XP machines locally, KY USA, and can only guess at Asian deployments. I suspect that XP will hang around much longer than other systems have. Just seems like lots of people don't want to let it die. – user4511 Feb 22 '17 at 4:20

I would say not yet. Mainly because XP marks the edge of its technology type which is still common with that of Windows 7. Games and applications which run fine on XP will still work correctly on 7-10 in most cases. That is not the case for WinME and its predecessors. This is especially true if they do any type of direct hardware access which precludes XP or NT's HAL.

I would hazard that XP would need to be in End of Life status for at least 6 years before anything could be considered retro.

Edit: Another thing which allowed Pre-XP OS's to reach "Retro" status earlier is the fact that there isn't a whole bunch of information on the Web currently for those operating systems. Where as XP and its newer iterations have existed in the time when the Web was prolific. As a result there are still numerous websites with current information reflecting current Hardware and current software.

  • 1
    Thanks for pointing out the "common technology base" with recent Windows versions - I think that is the fact that mostly disqualifies XP – tofro Feb 20 '17 at 14:30

Here's one way to look at it.

Windows 2.0 was released in 1987. That means 2.0 was about 14 years old win XP came out, and it almost certainly would have been considered "retro" at the time.

Windows XP is now over 15 years old.

  • 2
    Yes, but the Windows 2.0 market share was much lower when Windows XP came out than Windows XP is now. – JAL Feb 17 '17 at 19:50
  • 3
    When Windows XP came out, Windows 2.0 was a curiosity that would occasionally show up on abandonware sites. Today, you're not going to find XP on abandonware sites, you're going to find it on mainstream warez sites. – Mark Feb 17 '17 at 20:16
  • Windows XP shares much code with early versions of Windows NT (I'd hazard a guess that significant chunks are shared between XP and NT 4.0, and less with NT 3.x). Windows NT was originally slated to become OS/2 3.0 (NT 3.x and 4.0 could run OS/2 applications), but everything about that changed with the IBM/Microsoft split of 1989-1990. Windows 2.0 is part of the lineage that led up to Windows 3.x, Windows 9x and ended with the release of Windows ME; NT's lineage in this regard started at NT 3.1 and went through NT 4.0, Windows 2000, Windows XP and on with the versions of Windows we use today. – user Feb 26 '17 at 18:24
  • @Mark You're going to find GBC titles on mainstream warez sites as well, but GBC is certainly retro. – forest Mar 24 '18 at 21:39

Win XP still receives updates till 2019 (one only has to change a registry setting).

Can such software be considered retrocomputiong? I doubt...

  • 2
    Could you explain more? That's interesting – Cullub Mar 22 '17 at 14:26
  • @cullub google.com/… – Anixx Mar 22 '17 at 15:00

Windows XP as a whole? Certainly not yet. But 16-bit subsystems of Windows are strongly connected to the topic, and retrocomputing.SX should encourage questions about running 16-bit software on Windows (XP included). Therefore “doing under XP” may not be considered a definite off-topic and respective tag should be created (supplied with usage guidance).

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .