> The tradition goes back to SimCity, which only ran on Windows 95 because Microsoft wrote special code for it... Yes, Microsoft modified its OS just to cover up one game’s bug... The SimCity hack was a one-off
The stuff around SimCity in this article is poorly written. It seems the author has misunderstood it to have been a one-off special case. I looked back at Joel Spolsky's article, and he does make it quite clear that the SimCity example was just one of many: "it still ran old 16 bit software perfectly. Microsoft obsessed about this, spending a big chunk of change testing every old program they could find with Windows 95".
The article does later mention "Windows XP’s hidden database of app fixes" but doesn't appreciate that the concept really started with Windows 95, and in a big way, not "just to cover up one game’s bug".
Rough guess but this is all content farming of Raymond Chen posts at https://devblogs.microsoft.com/oldnewthing/ though I can't find a smoking gun. I definitely remember an article about Intuit doing something some dumb they had to keep an internal API around reporting the wrong value or something like that.
> On beta versions of Windows 95, SimCity wasn’t working in testing. Microsoft tracked down the bug and added specific code to Windows 95 that looks for SimCity. If it finds SimCity running, it runs the memory allocator in a special mode that doesn’t free memory right away. That’s the kind of obsession with backward compatibility that made people willing to upgrade to Windows 95.
The article’s content is generally quite weak, could probably have been covered in an old school tweet with screenshots and nothing of value would have been lost.
The only positive thing I can say about it is that it doesn’t appear to be written by AI.
I don’t think this person truly understands what they’re writing about, they did a little research, made some screenshots and put enough words in the article so that it will show up in search engines.
It’s reminiscent to me of many articles on Windows error codes, almost all of them are basically generic, generally quite long with drivel, and steering you to install some “fix it” software. Any technical person coming across these just finds them annoying as they clearly gamed the search results with unhelpful content.
You can see my other comment, but short version is that I think it’s just content for the sake of content without the person actually really understanding the topic much more than most of their readers.
It works fine. The processes get a fenced off piece of memory and are allowed to portion it out to themselves any way they prefer. Syscalls are designed to be mostly neutral about how the app wants to proceed.
On a system with an MMU, applications have their own address space that the allocator draws from; pages need to be centrally allocated, but individual allocations can be handled in userspace
are you implying the fact that they proactively threw in a custom heap manager for a game is supposed to be a bad thing? The fact that almost any 25+ year old executable runs on Windows to this day is fantastic, they're seemingly the only guys who take backwards compatibility seriously.
I don’t know much about CPU and memory back then, but did they not have some kind of memory access debugging mode where it would match every access against the alloc map, understanding that it could be very slow, but would find some of this kind of bug? Come to think of it, do we even have that today?
I was imagining more like some kind of primitive CPU mode that sends every single RAM request to a software exception. Something you would normally have turned off but available for truly heavy duty debugging.
> The tradition goes back to SimCity, which only ran on Windows 95 because Microsoft wrote special code for it... Yes, Microsoft modified its OS just to cover up one game’s bug... The SimCity hack was a one-off
The stuff around SimCity in this article is poorly written. It seems the author has misunderstood it to have been a one-off special case. I looked back at Joel Spolsky's article, and he does make it quite clear that the SimCity example was just one of many: "it still ran old 16 bit software perfectly. Microsoft obsessed about this, spending a big chunk of change testing every old program they could find with Windows 95".
The article does later mention "Windows XP’s hidden database of app fixes" but doesn't appreciate that the concept really started with Windows 95, and in a big way, not "just to cover up one game’s bug".
Rough guess but this is all content farming of Raymond Chen posts at https://devblogs.microsoft.com/oldnewthing/ though I can't find a smoking gun. I definitely remember an article about Intuit doing something some dumb they had to keep an internal API around reporting the wrong value or something like that.
From an article on this[0]:
> On beta versions of Windows 95, SimCity wasn’t working in testing. Microsoft tracked down the bug and added specific code to Windows 95 that looks for SimCity. If it finds SimCity running, it runs the memory allocator in a special mode that doesn’t free memory right away. That’s the kind of obsession with backward compatibility that made people willing to upgrade to Windows 95.
[0] https://www.joelonsoftware.com/2000/05/24/strategy-letter-ii...
The article’s content is generally quite weak, could probably have been covered in an old school tweet with screenshots and nothing of value would have been lost.
The only positive thing I can say about it is that it doesn’t appear to be written by AI.
I don’t think this person truly understands what they’re writing about, they did a little research, made some screenshots and put enough words in the article so that it will show up in search engines.
It’s reminiscent to me of many articles on Windows error codes, almost all of them are basically generic, generally quite long with drivel, and steering you to install some “fix it” software. Any technical person coming across these just finds them annoying as they clearly gamed the search results with unhelpful content.
One may even speculate the author got some, or a lot of ... "help", if it's even a real author.
Why show a screenshot of a different game (SimCity 2000) when discussing SimCity?
You can see my other comment, but short version is that I think it’s just content for the sake of content without the person actually really understanding the topic much more than most of their readers.
This is a great example of how implementation becomes the specification.
This is why memory allocators should be linked into applications and should not be considered operating system components.
Userland system DLLs need to allocate memory too.
And when you are running more than one application at the same time how is that supposed to work?
It works fine. The processes get a fenced off piece of memory and are allowed to portion it out to themselves any way they prefer. Syscalls are designed to be mostly neutral about how the app wants to proceed.
On a system with an MMU, applications have their own address space that the allocator draws from; pages need to be centrally allocated, but individual allocations can be handled in userspace
I wouldn't be surprised by windows components that use this. Wish I could run 16-bit apps tho.
> Wish I could run 16-bit apps tho.
https://github.com/otya128/winevdm
https://github.com/leecher1337/ntvdmx64
What else could be expected of Redmond's finest?
are you implying the fact that they proactively threw in a custom heap manager for a game is supposed to be a bad thing? The fact that almost any 25+ year old executable runs on Windows to this day is fantastic, they're seemingly the only guys who take backwards compatibility seriously.
A hidden z80 emulator...
I don’t know much about CPU and memory back then, but did they not have some kind of memory access debugging mode where it would match every access against the alloc map, understanding that it could be very slow, but would find some of this kind of bug? Come to think of it, do we even have that today?
This is what sanitizers do.
I know compilers have advanced, but I was wondering about a runtime debug check.
A sanitizer is that runtime debug check. The compiler just inserts those checks.
Okay. Then I have to assume that they didn't do it for SimCity! Thanks.
Sanitizers come with a lot of overhead, and the tech wasn't readily available in SimCity's day, certainly not for Windows 95.
I was imagining more like some kind of primitive CPU mode that sends every single RAM request to a software exception. Something you would normally have turned off but available for truly heavy duty debugging.