Of all the technical books I read, here are two that helped me greatly to learn technologies. “Thinking in Java” and “Thinking in Patterns” are two books by Bruce Eckel (http://mindviewinc.com) that opened my mind, not only by the technology that cuold be explained, but in the simple way the author has to explain things.
It is no accident the title of this post. I speak as an introduction to the concept of integration between IBM i and open technologies as an extension of what I wrote in my previous post.
When AS/400 changed its technology from CISC to RISC processors back in the ’90s, used the fact that its own operating system runs on the “Virtual Machine” Technology Independent Machine Interface (TIMI), and generated a series of automatic routines OS/400 to convert programs compiled in CISC to adapt to RISC: There was no need to reprogram anything! And only had to recompile the 5% of the code that Was working in the AS/400 of the moment. It was a masterful move by the giant blue And it is still today the change in technology more “clean” in the history of computing. Such was the impact of this, that was the “workhorse” from all the lectures of Frank Soltis Chief Engineer at IBM, responsible for the AS/ 400, who even went so far as to scorn of Intel, which made a change in technology of 32-bit to 64-bit so cumbersome, that today there are still thousands of commercial systems that still cannot make the conversion, because the cost is retyping completely.
When IBM gave account of the technological leap that should make its formerly called AS/400 to enter the competitiveness of the current world of the requirements Express, said “Hey! Why can’t we help our customers as we did before? Let us give them a tool to convert all your screens to green Web! “. As happens in the cinema, where the second parts are not as good as the first, this second attempt of IBM resulted in a true disaster.
What was the problem? In the first case, the development paradigm was intact, only change the processor technology. This means that the mode of the program does not change, nor the languages, or the screens. In the second case… the paradigm changes completely! There is no longer interactive programs, because the philosophy of the Web is that everything (incredibly) is Batch! There is no equivalent to a Display file in any side! And therefore there is no equivalent to the old familiar EXFMT of RPG! This was the first reason for which the tools of transformation of IBM (Web facing, and then its evolution HATS) failed miserably.
The second reason, is that the Web tends to be “functionally correct” (by this I do not mean to pretty colors and nice images, but to the organization of each button, text box or table, which is designed to be functional! If you do not agree, they come to see Facebook, Twitter, or any of the interfaces of Google! They are not beautiful … are comfortable and intuitive). Think this through for a second: How is it possible that a user interface of 80×24 characters, which only supports text, compete in appearance and functionality with an interface designed to fill the entire monitor in a resolution of 1024×768 pixels, which supports interactions in real time each time it is passed a Mouse above or by pressing a special key?
There is no case. There is no software tool can be inferred, that in a green screen is wanted to put a autocomplete function in a text field, or you want to see an explanatory table when you pass the mouse over a keyword, or you want to reorder fields in a subfile with a click of the Mouse… software tools do not think. Developers may do so. For this reason, is that no conversion tool of the Web that could propose IBM (or any of its competitors, of which there are many on the Internet) will be able to compete with the results of a Web development plan for the Web, and with the head start in the functionality.
This is without counting the number of database fields that, because of a lack of the appropriate data types in his time, took no sense in the database but in the logic of the programs. For example, this is a typical case of the date fields of the IBM API, that does not use a Date data type, but a Packed: Whoever can guess that a numeric field with the 1120205 content actually means “February 5,2012 “?
Then, if the conversion tools do not meet the expectations of senior management, the solution is to throw everything to the trash and start over again on another platform? Unfortunately, many companies thought that yes, either by disinformation, or stubbornness of IBM to sell hats. I propose another solution, which was tested with success in several places more “avant-garde”: Do not pull the AS/ 400… seize everything that they have there and integrate the same with the platforms that are designed for this type of functionality! I do not agree with the idea of converting to the specialists on the AS/400 to new technologies such as Java, HTML, and frameworks several, But I do think that the specialists on the AS/400 know the logic of the business and the customer is internal or external, therefore they should be the analysts, the “owners” of the data. Reality demonstrates to me that for the management of the requirements of user interface or connectivity with external systems, the best business that a company can do to stay technologically viable is to hire a specialist in these topics, And let the iSeries developer perform what should: Handle the core of the business.
My goal for the next few posts, will be to provide an introduction to the frameworks and methodologies of connectivity to be able to integrate the team of developers AS/400 with a team of developers of open technologies for power, between the two, having the best upgrade option of the technological infrastructure that have And at the same time receive less “why it takes so much time to make a requirement? “… or “I have to settle with a green screen if my competition has something else? “.