Re: Informix running on Linux
Doug McAllister said >The overhead for Informix to do a port to any platform and then to >maintain it is about the same regardless of whether it is for Linux, >Solaris, or what-have-you. If there is significant overhead in porting standard C using POSIX/Single UNIX Spec interfaces then they need to attend to their code base anyway... If they feel there is functionality missing from Linux that they *need* then they could start by actually telling someone what it is - *specifically*. >Now think about the return on investment for >that effort. Does any one think that the ENTIRE Linux market could equal >ONE large 'mainstream' deal? In simple up front value? Who knows? Define a "large" deal. Estimate the size of the "Linux market". Estimate the penetration. Decide whether to go for up front profit in cash or whether to develop a solid market base for the future. Do a risk analysis to determine whether all your eggs are safe in one basket... >Do you think that the shareholders of >Informix would stand for something like that? Also remember that >Informix is not a non-profit organization and they can't give something >away. Of course they can! Many companies "give away" some of their product in order to develop markets for their other products. You might not notice the loss leaders in your local supermarket but a couple of big computer companies spring to mind - Netscape and Microsoft. Companies frequently write off up front costs against expected future profitability. It's called investing in the future :-). >Anyone that is using Linux already proves that they won't pay for >software other than some 'token' fee. Rubbish! I know a *lot* of people that *bought* WordPerfect for SCO so they could run it on Linux. I know a *lot* of people that *bought* Informix for SCO so they could run their apps on Linux. As one MD of a *very* large firm said to me, "Linux is faster than SCO, cheaper than SCO, easier to support than SCO and runs SCO apps. Why should I buy SCO?" His one beef was that Informix couldn't or wouldn't provide development tools under Linux. Thus he was moving away from *Informix*. >Also, how many developers that are >developing on Linux and that want Informix are going to put their >product out for the same minimal profit percentage that they expect >Informix to take? Who can guess? It depends whether they hope to ship lots of units with a small profit on each or one or two units with a whopping profit on each. This is basic marketing. But there are other, equally important, issues to address in making the decision. >So lighten up and find some free database software to go with Linux. >When you decide to do some SERIOUS development, then maybe you will >spend the money for some serious software. They *are* going with low end Linux-native databases (not necessarily free either). These are people who are not learning Informix skills and no longer see Informix as an option. There is an outside chance that they might one day find a need to buy a large, expensive Informix engine - but they will scale what they have to the absolute limit first as always - and even then the absence of supporting tools would be a *big* minus for Informix. This is not a Linux or not Linux question either. I see SCO developers drifting from Informix and NT + SQL Server oozes into many of the breaches. Can Informix survive on a business strategy of selling a handful of units to a handful of large clients while the bottom end of the market crumbles away? I don't know, but the lack of market development is worrying. I'm reminded of the Bene Gesserit saying from Frank Herbert's Dune, "Here lies a toppled god, his fall was not a small one. We did but build his pedastle, a narrow and a tall one." Mike -- .----------------------------------------------------------------------. | Mike Jagdis | Internet: mailto:mike@roan.co.uk | | Roan Technology Ltd. | | | 54A Peach Street, Wokingham | Telephone: +44 118 989 0403 |