Re: please revive project 4GL++, using 4gl for Windows and New Era

Andrew Hamm wrote: >What have I missed? What's this EGL? Where's the early messages in this >thread? > Here it is: > Hi all! > > I think we should start a campaign to convince IBM to revive New Era > and Informix 4GL for Windows. Let me explain. > > Informix gave up on development tools long ago. Bad business, they > said. Unfortunatly, several years on this path, the "standard" tools > (VB, Java, etc.) are too "expensive" to work on. I mean: have you seen > how many lines of code you have to write in Perl, VB or Java to do > what in 4GL you could do with "SELECT col1 INTO var FROM table WHERE > col2=value"? Or the code necessary to create a small data entry form? > > IBM is announcing that 4GL is dead. The future, they say, is EGL. A > language nobody knows and that will not be source compatible with 4GL. > Partial migration tools will be available. Will they migrate the 4GL > programmers knowledge, too? > > I don't agree with this strategy. And I'm not alone. The 4GL instaled > base needs another solution. You may object that there already are > alternatives like 4J's, Querix or even Aubit 4GL, but they are not in > the right path (may be the last one is, but it's an OS project that > has very limited manpower). And they have not the "brand". > > What we need is a better 4GL! And we need it soon. And the best player > to do this is IBM. > > I think IBM would do this with not much trouble. It would require 2 > things: > > * Integrating the New Era syntax in the 4GL compiler (making it > 4gl++ as planned around the 90's) > * Creating a RDS runner that would run in Windows, like they did in > 4GL for Windows (just had to be prettier than that ugly thing they > released then!); a compiler/debugger in Windows would be nice too, > but not essential > > > IBM has the New Era compiler. They also have the 4GL for Windows GUI. > IBM should put a small team working in 4GL to incorporate those two > products in a new 4GL version. We wouldn't need anymore, by now. > > What IBM gains with this is: > > * Every user of 4GL will be willing to migrate to the new version > because it will be totally compatible (uprgade and maintenance $$!) > * 4GL user satisfaction will climb > * 4GL *will* be used for new projects, because it's much more of a > RAD tool than many other alternatives > > > The roadmap for this product, afterwards, should be: > > * Drop the native compiled 4gl version - it's too much trouble (have > you run benchmarks on rds vs compiled? i have, and gave up on > compiled 4GL long ago!) > * Generate Java Byte Code so that the only thing needed to run 4GL > would be a JVM > * Integrate 4GL++ in Eclipe > * Extend the syntax (yes, syntax, not API - it's a 4GL!) to support > XML, better UI, better data structures (dinamic arrays, > Collections, etc.), multiple connections to DB, Reflection, > Threads, Perform and ACE integration, File&Net I/O, etc. > * WYSIWYG forms > * Make Java classes usable within 4GL (like JPython) and 4GL > callable by Java programs > > > I think this is the way to keep 4GL users/developpers happy. Migration > tools exist to other platforms. If 4GL is forced to migrate, IBM will > be competing for a client it already has gained before! > > Please IBM, rethink your strategy! Technically, it's not so hard to do! > > TIA, > > Fernando > > PS: long live "enter2"! -- Fernando Fernandez MoreData - Sistemas de Informa''o, Lda. http://www.moredata.pt