Greg;
 
   There are some fundamental differences in the way that VistA looks at data and more traditional systems look at data.  The reaon that VistA has been so successful is that each new application has added color and aditional depth to the corporate data model for the applications being built and evolved.  So many of the new applications are actually orthogonal to the previous application.  The earlier applications are not necessarily tied to the later developed applications which are built upon them.   The new data structures don't necessarily need to interfer with the form and function of the original applications.  Are there dependencies established with the earlier application for the newer applications?  Perhaps, if the developer doesn't understand the full _expression_ of the applications they are using as base assumptions for their new model.  But at least they had that existing framework to hang their models on and that jumped their design from concept into a working model in short order.  Now that seems too short cut for good software development.  Possibly true, but it is good enough to get the model into the hands of the folks who have to live with it and use it every day.
 
  Now this provides a model which can go through many of the iteritive steps of development very quickly without having to go through the laborious and costly effort of capturing enough of the specification to build the next release (many months and man-hours away).  Not only that but this type of rapid prototyping can happen in a number of places at the same time.  The best solution frequently is some where in the midle of all of these development efforts.  And in some cases, the solution is only of value to the people of this one facility.
 
  The goal is to provide a mechanism which is orthogonal to the code enhancement path so that the localizations are not overwritten.  WorldVistA will be providing guidelines for how this can be done without fear of being overwritten by centrally released enhancements in the future.  If these enhancements can be brought back to WorldVistA, we will be working these enhancements into the main code base as standard release if they are considered to be useful and desired by the general population.
 
    Best wishes;   Chris
----- Original Message -----
Sent: Friday, July 07, 2006 11:07 AM
Subject: [Hardhats-members] Favor noninvasive methods

Reflecting a bit on this morning's meta-discussion, it occurred to me that I can sometimes cause some consternation among other VistA developers because I tend to introduce new routines and new files in preference to making incompatible changes to existing code. Of course, what I have in mind is not duplication of code, but rather a form of dependency injection. Why don't we see this more often? I think one reason is that we tend to manage file numbers and routine names as scarce resources. given the infrastructure we have in place today, maybe they are scarce resources, but I suppose it didn't occur to me that this is a big reason why we have so often relied on invasive methods (like "in place" code changes) to extend VistA. It's not just a matter of having "room to grow", but something more subtle. The very patterns that support extensibility without introducing tight coupling generally do require some supporting structures, and they need to live somewhere.


Gregory Woodhouse

"Judge a man by his questions not 
his answers."   --Voltaire




Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642


_______________________________________________
Hardhats-members mailing list
Hardhats-members@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/hardhats-members
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
Hardhats-members mailing list
Hardhats-members@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/hardhats-members

Reply via email to