3 ms·

The article has a sentence on the matter, but just to expand a bit more: Windows 7 is not a new piece of self-standing software. It includes past Windows releas

The article has a sentence on the matter, but just to expand a bit more: Windows 7 is not a new piece of self-standing software. It includes past Windows releases, and future Windows releases include it. It doesn't make sense for Microsoft to do this. Besides, I don't see any person/people being able to understand such a complex feat of engineering (and legacy) sufficiently to maintain it for years to come.

gmueckl 7y ago
Well, there's Wine and ReactOS as two major efforts to reimplement Win32 user space and the operating system as a whole. These are big team efforts, certainly, but they manifest a very detailed understanding of Windows, mostly obtained through reverse engineering. I would assume that there is enough interest in maintaining Windows - especially commercial interest - that finding and funding a sufficiently large and knowledgeable team is actually not unrealistic. It's just unrealistic to assume that the source code will ever be published willingly by Microsoft as long as any part of it is a direct ancestor to a product they still sell.
- grenoire 7y ago
  Good point about Wine and ReactOS; I would still think that there must be so much hidden away and swept under the million rugs. As for the commercial interest part, just thinking about how video games' source codes are so rarely published saddens me a great deal. Even after the commercial interest is lost and the IP is no longer in use, companies don't really think much about preserving the source code.
- krilly 7y ago
  You make a good point, although I think it's pretty clear that ReactOS cheated. https://www.theregister.co.uk/2019/07/03/reactos_windows_research_kernel_claim/ https://www.theregister.co.uk/2019/07/03/reactos_windows_res... What about open sourcing, say, windows 95? That seems more achievable, and probably more helpful.
  - gmueckl 7y ago
    Windows 95 is a kludge by today's standards and it's been the little brother of the more complete Win32 implementation in the NT line. As a result, it is missing a ton of things that are essential parts of modern Windows. I don't see any value in that.
    - ChrisSD 7y ago
      Arguably, the kludge is supporting Win32 on an NT kernel. The win32 layer on top of NT is a mess of backwards compatibility hacks with more modern NT and NTFS features hammered in. The native NT layer is much better but officially it's unstable. Compared to that, Windows 9x was less kludge and more a straightforward implementation of Win32, though of course lacking modern features.
  - wongarsu 7y ago
    > What about open sourcing, say, windows 95? That seems more achievable, and probably more helpful. The Windows NT4 source code leaked a while ago and can be found all over the internet. Of course that doesn't really help ReactOS because they officially can't use it, but that's probably the closest we will get to MS open sourcing Windows.
  - mehrdadn 7y ago
    I'm not sure it's "pretty clear"? See the entire thread: https://news.ycombinator.com/item?id=20343007 https://news.ycombinator.com/item?id=20343007

gmueckl 7y ago

Well, there's Wine and ReactOS as two major efforts to reimplement Win32 user space and the operating system as a whole. These are big team efforts, certainly, but they manifest a very detailed understanding of Windows, mostly obtained through reverse engineering. I would assume that there is enough interest in maintaining Windows - especially commercial interest - that finding and funding a sufficiently large and knowledgeable team is actually not unrealistic. It's just unrealistic to assume that the source code will ever be published willingly by Microsoft as long as any part of it is a direct ancestor to a product they still sell.

grenoire 7y ago
Good point about Wine and ReactOS; I would still think that there must be so much hidden away and swept under the million rugs. As for the commercial interest part, just thinking about how video games' source codes are so rarely published saddens me a great deal. Even after the commercial interest is lost and the IP is no longer in use, companies don't really think much about preserving the source code.
krilly 7y ago
You make a good point, although I think it's pretty clear that ReactOS cheated. https://www.theregister.co.uk/2019/07/03/reactos_windows_research_kernel_claim/ https://www.theregister.co.uk/2019/07/03/reactos_windows_res... What about open sourcing, say, windows 95? That seems more achievable, and probably more helpful.
- gmueckl 7y ago
  Windows 95 is a kludge by today's standards and it's been the little brother of the more complete Win32 implementation in the NT line. As a result, it is missing a ton of things that are essential parts of modern Windows. I don't see any value in that.
  - ChrisSD 7y ago
    Arguably, the kludge is supporting Win32 on an NT kernel. The win32 layer on top of NT is a mess of backwards compatibility hacks with more modern NT and NTFS features hammered in. The native NT layer is much better but officially it's unstable. Compared to that, Windows 9x was less kludge and more a straightforward implementation of Win32, though of course lacking modern features.
- wongarsu 7y ago
  > What about open sourcing, say, windows 95? That seems more achievable, and probably more helpful. The Windows NT4 source code leaked a while ago and can be found all over the internet. Of course that doesn't really help ReactOS because they officially can't use it, but that's probably the closest we will get to MS open sourcing Windows.
- mehrdadn 7y ago
  I'm not sure it's "pretty clear"? See the entire thread: https://news.ycombinator.com/item?id=20343007 https://news.ycombinator.com/item?id=20343007