An interesting piece of news hit the press about two weeks ago regarding the NYSE's announcement to to move to Linux with what HP owning the deal. Even though there isn't a whole lot of details publicized about this move it appears that the move is taking place within the HP account (not a win from a competing vendor) - the platform is being transitioned from what appears to be a mix of HP-UX and NonStop. In other words it is most likely a management-driven (read non-technical) change based purely on costs. Which to me raises a very interesting question regarding the reliability of the upcoming Linux based platform. Yes, I'm firmly convinced that Linux is still not up-to-scratch compared with the mission critical longstays like HP-UX, AIX and Solaris and chances are that NYSE has some headaches in front of them. But even that is not my biggest concern regarding this move, it is rather the *how* HP will deliver support for such a high stakes platform while not really owning any Linux IP. I know the Linux is open source, blah, blah, blah, and anyone can theoretically provide support for it. But that is theory. In practice, unless you have staff engineers routinely maintaining the code and you have direct control over those engineers, only then you can deliver sufficient support for highly demanding SLA environments. Well, HP hardly has any Linux developers on staff and HP has been avoiding every opportunity to make any serious software play with Linux sticking with selling hardware for Linux and relying on third parties to provide the Linux real support. I can only attest to the fact how bad RedHat support can get simply because they don't have developers that cover the entire breadth of the product they're selling and a lot of times they simply resort to posting questions to open forums as a way of helping the customer resolve their issues. Most definitely that wouldn't fly in case of New York Stock Exchange, those guys will want fixes "now" as soon any problem will start getting anywhere near serious. So I'm not sure how well this Linux-on-HP arrangement would work with NYSE when an integrated solution owned by one vendor (hardware + OS = HP-UX) would be more appropriate. On the other hand I can easily see how it could have happened, HP was most likely given one of two options: get lost or deliver a cheaper solution. It looks like HP chose the latter and I guess I wouldn't be surprised to see the headlines that the biggest stock exchange in the world had issues that couldn't be resolved on time... I'm saying that Linux is bad or unreliable, but don't send a body to do a man's job...