There has been a lot of talk from PSC in the last couple of years about their Application Transformation Approach product from their consulting group and I have seen presentations on a number of success stories. I believe this is also on the agenda for Exchange. This makes me wonder about what kind of experience there is with application transformation projects out there, so I thought I would do a little survey. Please respond either publicly or privately as you prefer and if I get more than a couple of responses I will attempt to synthesize and post.
For these purposes, "application transformation" will be considered to be any project whose focus is a significant architectural change in the current application. This includes, among other things:
separating UI from BL
moving to a new UI, often in association with separating the UI
implementing SOA/ESB
working toward OERA compliance
separating out a data access layer, possibly in association with supporting a new database.
But, please include any project which you think fulfills the characteristic of a significant architectural change, whether it is on this list or not. To maximize the responses, I am interested in input not only from those who are part of the company making the change, but also those participating in a particular component as a consultant or even just something you know about, but have no direct involvement.
Questions:
1. What is your relationship to the project? Principal/Architect, developer, consultant, witness?
2. What was the nature of the transform being done?
3. At what stage is the project? Twinkle in the eye, planning, started, mid way, near completion, done, abandoned?
4. Did PSC consulting participate in setting the directions? If not, who did?
5. Is the company doing the transformation an AP or an end-user? Is the application originally written in-house or purchased?
6. What version of Progress is the target?
7. Is any use of OO involved? Imitation OO?
8. What is/was the driving force behind the transformation? New UI, interconnectability, distributed deployment, maintainability, aging architecture, move to OO, move to OERA, competitive pressure, need for web access?
9. What was done/is being done well? What was done/is being done poorly?
10. Was any form of modeling used as a part of the process?
11. What tools were used? What tools would have helped had they been used? What tools would have helped if they had existed?
12. Any words of wisdom for those considering similar projects?
Questions:
13. are you considering another development environment?
14. are you creating a new front end in .Net, if so, why?
I wonder if you get feedback on these mission critical questions
I suppose that I consider the new development environment question to be one that either parallels or precedes a consideration of application transformation. I.e., it is more either or.
The new front end in .NET to me is subordinate to the questions I asked. I.e., if one had already restructured to separate BL and UI using Java or web or ABL clients and then went to add or move to .NET clients, that is not really application transformation because the architecture is already there. But, if one is implementing a .NET UI and thus having to restructure the application to separate UI and BL to make that clean, then the significant factor in the current context is the shift in architecture, not the particular language used for the client.
And no, there has not yet been one response.
Possibly this is because most of the application transformation projects have been done by partners and most of them seem to be relatively invisible on PSDN or any other of the on-line forums. Yours truly being a notable exception, of course.