You make a valid point. There are a lot of crazy people out there, running every kind of OS on every kind of hardware for every kind of company. And any untested system, no matter what, is not ready for the enterprise.
But there are a great number of large companies running Win2K Enterprise or SuSE on mission critial applications. I've worked on both NASDAQ, DOD and CIGNA, all of whom considered their missions critical, and we've successfully deployed a MS solutions. MS itself seems to be happy with BSD running HotMail. I'm working on an enterprise SuSE contract now far a large electronics company, and I'm surrounded by QA.
Its not true to say that there has not been unit testing There are ISO9000-certified IT structures using these systems. And there are well-known stable GNU apps, like Apache. There is a lack of published specs, but they're usually skewed to benchmarking and/or are performed on hardware you don't have. You need internal testing, and you don't have access to those figures to state whether or not they are happening.
You sound like an AS/400 veteran, in which case your comments make sense. An AS/400 is more stable than granite. But I've seen them brought to their knees by things like CASE tool conversions gone bad.
As you said, testing is paramount. But you needn't assume that there is no testing going on because you haven't seen it.
Re:Crazy people.
Posted by: DCallaghan on June 14, 2002 09:42 PMBut there are a great number of large companies running Win2K Enterprise or SuSE on mission critial applications. I've worked on both NASDAQ, DOD and CIGNA, all of whom considered their missions critical, and we've successfully deployed a MS solutions. MS itself seems to be happy with BSD running HotMail. I'm working on an enterprise SuSE contract now far a large electronics company, and I'm surrounded by QA.
Its not true to say that there has not been unit testing There are ISO9000-certified IT structures using these systems. And there are well-known stable GNU apps, like Apache. There is a lack of published specs, but they're usually skewed to benchmarking and/or are performed on hardware you don't have. You need internal testing, and you don't have access to those figures to state whether or not they are happening.
You sound like an AS/400 veteran, in which case your comments make sense. An AS/400 is more stable than granite. But I've seen them brought to their knees by things like CASE tool conversions gone bad.
As you said, testing is paramount. But you needn't assume that there is no testing going on because you haven't seen it.
#