Should we be using NewEra

I would like to bemoan the pain we had experienced with NewEra and to get some advice from you experts out there whether it is worthwhile to continue with it (and Informix) in the light of the uncertainty that seems to hangs over the future of NewEra. Several years ago we used Informix 4GL and had reasonable success with it. The database was stable, fast and we liked it. Then we had to change to a GUI environment. We are operating in a scientific environment and with the advent of GIS, it was required that all new applications must be graphical. We must develop a number of very important mission critical applications, which will happen over a number of years. After several products were investigated, we decided on NewEra (Version 1). We realised that the product was, at that stage, still immature, but supported OO development, which put it in a class of its own in comparison with what other DB vendors could offer. We also had a look at third party front-end tools, but the combination was expensive and we experienced incompatibilities when it was tested. We experienced lots pain to implement NewEra. The first problem was support. Apart from the relative elementary NewEra training, we had to sort out most of our own problems. With Version 2 we stepped into the trap of the Application Framework Class Library. The unsatisfactory interface and bugs wasted much time as we tried to devise workarounds. Most experts with whom I had discussions lately, strongly advise users not to use it and rather to develop your on base classes. The lack of basic easy-to-use tools, such as tabwindows, was also a time waster - we had to develop or find such aids. After the struggle with the long and steep learning curve, little support and building our own base classes, it seemed as if our problems were easing off. The release of the much-improved Version 3.0 and now Version 3.1 with OCX support, also proofed that the product has reached maturity. No such luck! The news of the possible end of NewEra, really put the provable cat amongst the pigeons. The biggest Achilles heel is the lack of NewEra programmers and the long time it takes to train them. We are dependent on contractors and they are now extremely reluctant to invest more time in mastering a dying product. We had many discussions about possible alternatives, but came to the same conclusion as Nils Myklebust. Here is an extract from his reaction: A significant problem for Jeff Hudson and Informix is that there aren't really that many development tools out there that can be used for major (Informix?) database developments. He mentions Visual Basic. Well it's nice for a small client only applications. When it comes to major applications that also need server side programming it's not a viable solution unless you decide up front that Windows NT shall forever be your only server option. He mentions Java. I happen to believe that Java is and will be a tool for major developments. There are now several large, to very large, systems being developed in Java, and it all looks good. There is however still troubled water ahead, and it will still be some time before it's tested enough for most purposes a database development tool is needed. He mentions no other. The reason is simply there are no other major tools for the great public. Yes there is Delphi, but essentially with the same limitations as Visual Basic (although it's clearly a better language). PowerBuilder has the same problems. Then there is Forte. Too expensive, coming from a small company that feels the competition from Java. Then there is a plethora of tools from companies all over the world. All are relatively small, however, generally very expensive, and not an option for many. NewEra was actually always cheaper than most of these, and now with better functionality than most. Last there is C or C++. In the database development arena this is a loosing proposition. If you develop an OS, a database engine or even a word processor or similar tools there is hardly any way around C. It simply take s too long to create finished database type applications with these alone. Many excellent database programmers would hardly be able to start at all. Now here is the essence of my dilemma: We have to develop a mayor application, which will take several years, and we need a reliable supported development tool which will be maintained and enhanced for a number of years. We are not in the research business. Our vendor must safe guard our investment by providing transition paths for new emerging technologies. Peace of mind is very essential part of the deal. For us using the best database without a proper toolkit is like eating gourmet soup without a spoon ' not enjoyable! I am afraid that we might have reach the end of a road. Am I overreacting, or is it time to change vendors before investing more money in a mayor project which can potentially affect the life of many people?