Re: IIUG - Linux and Informix

Lets sum up a little, do some numbers games to show the dollars in it for Informix and add some other arguments: 1. This is about supporting existing developers who allready have finished applications written in 4GL or possibly ESQL/C. Upgrades and further sales can and will happen reasonably fast - key point for when the dollars will arrive. 2. We are talking about small developers but large volumes. 3. In addition to 1. above we enable a number of new developers to use Informix. Some will need the engine in particular immediately, some will be more long term. The numbers games: I have heard some time ago that Informix have sold something like 250.000 copies of 4GL. I have estimated a price of US$ 500 per developer here. May be it should be US$ 300 with the SE engine and US$ 600 with OnLine with some split between them. Informix should probably do a little reasearch here. With these assumptions we can do as follows: Percent of total sold Total sold 4GL 250.000 Small developer 125.000 50 % Still using 4GL 62.500 50 % Will upgrade 12.500 20 % Price 500 US$ Upgrade 6.250.000 This might be the income the first year or so from direct upgrades/new sales of the new port to existing customers. Some of these have applications they sell. Some of these will be moderately successfull at selling. We can then estimate: Have some success: 125 1 % of "Will upgrade" No. sold 500 US$ to Informix 500 US$ success 31.250.000 It might take two years for this to materialize. It's likely some will become very successfull. Another estimate can be done: Become big: 12 0,1 % of "Will upgrade" No. sold 10.000 US$ to Informix 3.000 US$ big 360.000.000 This might materialize over the next two to four years. The US$ 3.000 to Informix here is due to larger installations, many on larger Unix machines. US$ total to Informix 397.500.000 This should be an interesting additional income to Informix even though it might be spread out over some years. All of the above are of course guesses and estimates only, and can be wildly wrong. They can however be as wrong on the downside as they can on the upside - meaning the numbers might be much larger as well as they can be smaller. The above are from sales related to current 4GL developers only. Remember that many will upgrade to new development tools, so the end user applications may not allways be in 4GL. The important issue for Informix right now is however engine sales. There will be new sales as well to new applications. Uses related to the Internet in general and the Web in particular comes first to mind. However I think there will be a pletora of other uses as well when you look at what Linux is actually used for. It's very hard to estimate sales for this segment so I will leave that to others. Here are some additional benefits to Informix that may be as important in a somewhat longer perspective as the dollars directly earned due to significant indirect sales resulting from this: 1. Retain the small developers and keep them as happy customers. If Informix does nothing they will loose a large number of these. They will not only loose these, but have a large number of disgruntled users, many of which will tell others of their experiences. It will become that much harder for Informix to penetrate other markets, particularly the Windows NT market that they are so interested in. 2. Exposure to Informix among developers and other computer users. This will be continued exposure of course, instead of the opposite as mentioned in 1 above. 3. A significant talent pool for others to draw on for consulting work as well as for emploiees. 4. A way for 4GL developers to move at their own pace to MS Windows or Web (often Java) programming. Without taking care of these developers for another few years using 4GL they simply can't afford to move to newer development environements. They will go out of bussiness, typically finding other work not related to Informix. 5. A way for new users to start using Informix on a relatively low budget developing new applications. These will probably be MS Windows or Web based (many using Java). The problems: 1. Technical problems of porting: This can't be true. There can't be any realy significant ones. If there are they should be possible to solve. 2. The "need" for an operating system developer to complain to about problems: a. Even Informix can't realy complain to Microsoft regarding Windows NT and have them fix anything. Windows NT is supplied "as is" with every realeas. This works well for Linux as well. Informix ports to some version, makes that work and users will have to use this version only. Obviously defects may be found that are caused by some defect in the selected version of Linux. What does Informix do in the Windows NT case? They probably fix it as best as they can trying to program arround it. They obviously report it to Microsoft. The result of this report? If they are lucky it gets included in a service pack. Otherwise may be it gets fixed in the next release. It may also easily go unfixed for years or a totaly new "solution" from Microsoft will have to be used by Informix. It may even be that I am wrong, and Microsoft fixes things for Informix. The whole issue isn't much different in the Linux field however. The problem for Informix is they do not have a single company they can send the defect report to. They can not point to a company and say: "It's their fault, not ours. You have to go to that company and have them fix their defective software befor anything can be done from our side." To be able to say something like this is important, but I venture to say that it's by no means a show stopper. Informix will instead have to simply say: "It's a defect in Linux. We have told this to some key people in the Linux community, but it's up to you as a customer to follow this up in any which way you can. Here is a detailed description of the defect we have run up against: ....." I can see no big difference between these two. Actually for me as a customer I might be in a better situation in the Linux market than if I use some of the other OS's Informix have ported to. But that is *not* important. The key issue is that the leagal as well as the customer support situation for Informix isn't much different. b. Informix can probably go to one of the commersial suppliers of Linux and get a support situation that isn't much different from what they get now from their other porting partners. Some other partners may have more people and resources that they can and will put on the porting effort, but with the