3 ms·

The issue is at that time updates for most shrink wrapped software were almost non existent and distribution even worse. Windows had an update facility and I be

The issue is at that time updates for most shrink wrapped software were almost non existent and distribution even worse. Windows had an update facility and I bet you sim city for dos did not. So, either windows yielded and patched it’s behavior to run sim city or sim city users couldn’t run it under windows full stop until maxis distributed a new distribution on media. In todays world it looks like a ludicrous solution to the problem. The game vendor needs to HTFU and distribute a patch via steam or whatever and that’s a totally reasonable stance for microsoft or apple to take, because the infrastructure for patch distribution is pervasive and robust now.

ukuina 3y ago
People forget how every software developer had to spin up their own patch delivery infrastructure.
- danbruc 3y ago
  If the company even existed when a Windows update caused an issue. And given the number of Windows installations, even the most obscure software or setup is likely to affect a lot of people.
- fnordpiglet 3y ago
  Back in the day it wasn’t even reasonably possible. You got a floppy/CD from the store and that was the end of the story. I mean, this is “Netscape just came out” times.
  - bluefirebrand 3y ago
    Yep. At best you had an updated version on the next run, but even that was unlikely.
    - brewtide 3y ago
      My AOL updates in the mailbox always lagged behind, week after week...
  - xoxxala 3y ago
    In the early 90s, we had a disk replacement policy and would mail people updated floppies on request (and a $10 S&H fee). Few people took advantage of it. We then started hosting some patches on Compuserve, GEnie and Prodigy.
dotnet00 3y ago
And it's worth emphasizing that Microsoft was/is incentivized to do this because especially with popular software, such bugs have a chance of resulting in users blaming Windows even if in reality it's the gamedev's fault. Similarly to how GPU vendors are incentivized to patch their drivers to fix bugs in a popular game release because the bugs might be blamed on the vendor instead of the dev.
- pseudosavant 3y ago
  I don't have the reference for it here, but in the first year Windows Vista was out >50% of crashes were due to buggy nVidia drivers. Microsoft assumed (incorrectly) that their ecosystem would get their shit together automatically. That nVidia would make solid drivers for the new Windows WDDM driver model. The year before Windows 7 came out I was working at a company (DivX ;-) making Windows software. We were getting contacted by different testing groups at Microsoft constantly. Some weeks three different people might contact me. Somehow they found my phone number? It didn't seem very efficient, but it was clear that they were allocating huge resources to this. They found unbelievably nuanced bugs (their QA was better than ours...). They wanted to know when we were going to have a fix by. They wanted to make sure they didn't have a Vista repeat. Vista SP1 was actually quite stable, but it was too late for the Vista brand already. With Windows 7 it seemed clear that the thing they cared about was: the software their users actually use still works after the update. Right or wrong, it was very user centric, because what user likes for their software to break? Nobody cares why.
  - lamontcg 3y ago
    Microsoft got to burn those kinds of resources on the compatibility problem due to having had a near-monopoly on the desktop back in those days. That isn't going to be repeatable for pretty much any other software company other than a FAANG, and certainly not of open source projects, not even Linux. People don't pay enough to open source for that kind of support.
    - gbin 3y ago
      It is true for the kernel though. Linus has a strict non regression (of the user code) rule. If some user code starts to break because of a kernel change, the kernel team takes the blame no question asked.

ukuina 3y ago

People forget how every software developer had to spin up their own patch delivery infrastructure.

danbruc 3y ago
If the company even existed when a Windows update caused an issue. And given the number of Windows installations, even the most obscure software or setup is likely to affect a lot of people.
fnordpiglet 3y ago
Back in the day it wasn’t even reasonably possible. You got a floppy/CD from the store and that was the end of the story. I mean, this is “Netscape just came out” times.
- bluefirebrand 3y ago
  Yep. At best you had an updated version on the next run, but even that was unlikely.
  - brewtide 3y ago
    My AOL updates in the mailbox always lagged behind, week after week...
- xoxxala 3y ago
  In the early 90s, we had a disk replacement policy and would mail people updated floppies on request (and a $10 S&H fee). Few people took advantage of it. We then started hosting some patches on Compuserve, GEnie and Prodigy.

dotnet00 3y ago

And it's worth emphasizing that Microsoft was/is incentivized to do this because especially with popular software, such bugs have a chance of resulting in users blaming Windows even if in reality it's the gamedev's fault. Similarly to how GPU vendors are incentivized to patch their drivers to fix bugs in a popular game release because the bugs might be blamed on the vendor instead of the dev.

pseudosavant 3y ago
I don't have the reference for it here, but in the first year Windows Vista was out >50% of crashes were due to buggy nVidia drivers. Microsoft assumed (incorrectly) that their ecosystem would get their shit together automatically. That nVidia would make solid drivers for the new Windows WDDM driver model. The year before Windows 7 came out I was working at a company (DivX ;-) making Windows software. We were getting contacted by different testing groups at Microsoft constantly. Some weeks three different people might contact me. Somehow they found my phone number? It didn't seem very efficient, but it was clear that they were allocating huge resources to this. They found unbelievably nuanced bugs (their QA was better than ours...). They wanted to know when we were going to have a fix by. They wanted to make sure they didn't have a Vista repeat. Vista SP1 was actually quite stable, but it was too late for the Vista brand already. With Windows 7 it seemed clear that the thing they cared about was: the software their users actually use still works after the update. Right or wrong, it was very user centric, because what user likes for their software to break? Nobody cares why.
- lamontcg 3y ago
  Microsoft got to burn those kinds of resources on the compatibility problem due to having had a near-monopoly on the desktop back in those days. That isn't going to be repeatable for pretty much any other software company other than a FAANG, and certainly not of open source projects, not even Linux. People don't pay enough to open source for that kind of support.
  - gbin 3y ago
    It is true for the kernel though. Linus has a strict non regression (of the user code) rule. If some user code starts to break because of a kernel change, the kernel team takes the blame no question asked.