Re: Informix running on Linux

David Williams wrote: > > In article <335F0B33.D712D8DF@mindspring.com>, Tim Schaefer > <tschaefe@mindspring.com> writes > >Besides if you really want to sell Universal Server, let go of 7.x. > No - 7.x handlesd purely relational data very well - Universal Server > will almost certainly be slower than Univseral Server because the > size of the working set of each function will be larger as there will > be > if (relational data) > ..... > else > ..... > endif > > somewhere in the code (or a case <database> statement) which will > result in some functions getting larger i.e. running slower. > David, I appreciate you wanting to defend Informix. This is a sign of extreme loyalty, and having read your posts for some time David, I feel that Informix had to earn your trust by proving the product to you. I doubt you accepted it hook line and sinker. Informix won your loyalty over by being worthy. Now the table is turned. We're being asked to stay loyal in the face of getting NOTHING for our loyalty. This is bad business thinking, thinking that comes from people who either don't know how to build business relationships, or forgot how it works. Linux is winning people over by realizing that operating systems in and of themselves are worthless. Without software running in the operating system, the computer is meaningless. Informix has got to realize that yes, the gravy is for the taking at the top, but the actual people who use it can ill afford to pay for it. These are the strongest advocates for the sale, not a bunch of hacks selling it, by taking the IS management to lunch, wooing them with kisses in the air, and promising them the moon. Sales people are not where the focus should be. So, having breathed all that, I say don't be so quick to defend Informix and say "no" to components of the port. Port the whole damn thing. All of it. Reward my loyalty by dumping in my lap everything I need to use the product on Linux, just as if it were on the HP or Solaris system at work. I'm not suggesting that everyone who uses it wants to buy it, but developers have traditionally brought a software product forward in the market place. They ultimately build the sales through consensus. We are being rewarded for our years of loyalty with a slap in the face. Informix has never contacted me from their sales department, and said, "Geez, Tim, we really like what you've done, promoting our product.". I <have> received software on my doorstep from another vendor, with a thank-you for merely mentioning their name in another newsgroup ONE time. This is what I'm talking about. Getting an affordable OS is just getting us started. Now we need software! Is this asking too much? I think not. Or how about the small business that wants to automate their business, but will never grow beyond a certain size. Linux makes that dream a reality. But it could also make it possible for a business to start out on x86 Linux, and then migrate to a bigger hardware platform. Hey, they make more money then buy a Digital system. The migration is painless because the OS is scaled. And what of an Open UNIX, that is not bound by vendor prejudices? This is new ground, and the WHOLE port should be a part of the deal. I want it all. I've earned it. > > I think they should > > a) dump SE it's outdated > b) stop porting Online 5.x and dump it > > Move everyone to at least OWS and DSA. Use the money from moving > everyone to DSA and the savings from not having to support > SE/5.x products to fund Universal Server and XPS. > Why limit the port? You can download the source code to the kernel. This means Informix could theoretically develop a Linux-Informix kernel that is optimal for them, and the products. Hell, with the source code you have NO limitations. And keep SE. As others have said it's still appropriate. But imagine looking at 7.x the way we look at SE. In five years you will. Let's get creative with this. > Evetually (when hardware has gone a generation) move everyone to > Universal Server and all 'big' sites to XPS (another generation of > hardware should be most large sites run >10 CPUS). > Scalability of an OS is only possible through a cooperative effort. Digital has been consistent with taking relatively new Operating Systems, such as NT, and making a success of it on their systems. You could see this happen with Informix if they could catch the vision. Being a leader involves a certain amount of following, but it also means not waiting for others to make your decisions, such as your hardware partners who apparently have you in their pockets. Operating Systems are now a commodity, and the coffin for UNIX has been rolled back to the storage shed. Linux is proof that UNIX can stay inexpensive. There should be no reason not to port Informix software to it. And there should be no reason why developers shouldn't get a complete copy of the software, to use on their Linux boxes at absolutely NO CHARGE to them. They might just be able to sell it for you. You can sell support as you currently do. Folks who don't want/need the support don't get it unless they pay for it. Hey let's really raise the stakes: Put Universal Server on Linux and blow everybody away! > >If you have more people using 7.x, they can migrate their experience to > >Universal Server, thus more momentum in the market. Who knows, if 7.x > >were on Linux, some of the Linux community might build some really > >cool tools for it. They're a busy bunch... > > > > Treu - I have starting writing my tools to analyses sqexplain.outs > and I'm also looks putting together sql to run against sysmaster for > my ondiags (Online Diagnostics) tools. > I eagerly await your tools and anyone else's out there. Build a web site. Start an ezine. Get inspired. Now is the time. Cheers, Tim > > -- > David Williams -- Tim Schaefer \\\\|// tschaefe@mindspring.com (6 6) ------------------------oOOo---( )---o00o--------- http:www.mindspring.com/~tschaefe