One of HP’s biggest faults is their overwhelming and excessive investment in overseas labor. You have >90% chance of getting someone in India and >95% chance of just getting someone overseas when you call for support. They have very few infrastructure engineers in the US anymore, and that saved them money initially; however, as time has consistently played out since the late 90s, outsourcing critical infrastructure personnel such as your engineers is 100% foolhardy.
I used to be one of their biggest champions, esp. for their hardware. I supported their ProLiant-class server hardware from the old Compaq days all the way through the latest generations. The platform is solid, but they’re not innovating. They’re stagnant, and companies like Cisco are designing integrated hardware platforms now (i.e. UCS) that provide packaged-deal pricing for much less than HP’s CTO or package prices provide.
HP will go down in history as a lesson to all future corporate bigwigs: try to overwhelm or silence your competition, try to buy them out and scuttle their brand through attrition, and the market will respond by out-innovating you.
*coughMicrosoftcough*
One good example of their ineptness is leaving OpenVMS virtually fallow.
As security becomes more and more important (and a bigger issue) having an OS with an architecture built with security (and reliability) in mind will become more and more important.
The first company that puts out a fully formally verified OS is probably going to get a *GREAT* headway in the market.