The Red Hat IPO stands out from the rest of the dot.com mania; unlike those companies with infinite or negative P/E ratios, Red Hat is making money today shipping a popular product. I've seen their shrink-wrapped distros not just in computer stores like CompUSA but in all kinds of mass-market stores like Best Buy for a year now, and they wouldn't still be on the shelf if they weren't selling well.
I've bought two shrink-wrapped Red Hat distros plus a Slackware and a SuSE. But now that I have a cable modem, I downloaded Red Hat 6.0 direct from Red Hat's ftp site (OK, actually from a mirror site) for no charge at all. I got the entire contents of their install and source CDs, but I don't get installation support. The Red Hat installer and Red Hat's RPM format for installation files are also GPL'd, so Martin Schiller is incorrect when he says "If they put together the installers...first, they've got the market." (Unless they write a new, proprietary installer.)
It would be legal for me to burn copies of Red Hat 6.0 with my CD recorder and mail them to you. Also, if I have a dozen PCs and I want to put NT on all of them, I have to buy a dozen NT licenses; but with my one copy of Red Hat 6.0 I'm free to install it on every machine in sight. Red Hat could change their licensing terms, but if they did all their customers would dump them overnight and buy SuSE, Debian, Caldera or Slackware distros instead.
t byfield repeats the well-known quote that "Linux is only free if your time is worth nothing." That's a nice piece of MS propaganda; does one learn how to use a commercial product like NT Server, or Novell Netware, or Solaris, instantly? As a desktop OS Linux requires maybe a hundred times more effort to get up to speed than the Mac OS (which is the easiest-to-learn OS I've ever seen; Windows is far from being so easy or intuitive). But for server tasks, which is where the Linux market is today, you need to learn so much about things like TCP/IP, that the difference in effort between NT with GUI configuration tools and editors, and Linux with vi and .rc files, amounts to practically nil. In fact, because it is documented so much better, Linux is easier to configure than NT or Netware. Nor can you just go by the number of installations - MS gets to keep the desktop OS market, where an OEM copy of NT Workstation costs about $100; but Linux is grabbing MS's customers for NT Server, at $450 an OEM copy, plus $25 per "client access license".
I'd sure like to have bought shares of Red Hat at the insiders's price of $14, but I am not an insider. As anyone could have guessed, it was up to over $40 a share by noon. The management of Red Hat, who either got to buy a good deal of those shares, or got stock options at $14, balanced what was best for the company against what was best for themselves personally.
Morgan, Stanley gets the big bucks in return for taking big risks, but unlike your ordinary "visionary" dot.com IPO, buying Red Hat was about as much of a risk for the initial buyers as getting out of bed. Ostensibly the purpose of an IPO is to raise money for the company. So couldn't they have asked, say, $21 a share instead of $14? If you're the CEO of Red Hat, your shares and stock options would only have doubled in value by noon, rather than tripled, but on the other hand it would also mean that the Red Hat company would have raised half again more cash from their IPO.
The economics pros on this list are probably used to this, however I am but a humble nerd and it puzzles me, why Red Hat's management allowed two-thirds of the money raised by their IPO go the early-investing insiders and only kept one-third for the company.
Yours WDK - WKiernan at concentric.net