I never said it was "nice" of Apple to offline the automatic update server for 10.3 and earlier. What I meant by what I did say was that it probably wasn't something they did specifically with a mean spirit, or out of spite.
My guess is that whatever system was hosting these updates is far from "modern" (it may not even have been new in 2003) and Apple isn't known as the kind of company that wants to keep a bunch of old hardware (and staff that knows how to run it) around. It was probably a fairly big decision for those involved, and there were probably cost-benefit analyses for various options.
As somebody who likes looking at and thinking about systems and infrastructure, and the business impact of those things, I would actually really like seeing information on what was happening. My guess is that it involved very big and high-end hardware from Sun, and a custom-written application that's expensive to port forward from Solaris 7/8. Couple that with the fact that a sufficiently big Sun system takes a veritable boatload of electricity, and the fact that disks for those older systems aren't exactly getting cheaper, and the decision is pretty clear to me.
It would also be interesting to see if they consciously decided to turn off a working system (not unprecedented) or if they came to work, found it dead, and decided at that point to discontinue the service.
My guess is that whatever system was hosting these updates is far from "modern" (it may not even have been new in 2003) and Apple isn't known as the kind of company that wants to keep a bunch of old hardware (and staff that knows how to run it) around. It was probably a fairly big decision for those involved, and there were probably cost-benefit analyses for various options.
As somebody who likes looking at and thinking about systems and infrastructure, and the business impact of those things, I would actually really like seeing information on what was happening. My guess is that it involved very big and high-end hardware from Sun, and a custom-written application that's expensive to port forward from Solaris 7/8. Couple that with the fact that a sufficiently big Sun system takes a veritable boatload of electricity, and the fact that disks for those older systems aren't exactly getting cheaper, and the decision is pretty clear to me.
It would also be interesting to see if they consciously decided to turn off a working system (not unprecedented) or if they came to work, found it dead, and decided at that point to discontinue the service.