>Have you been involved in any software port projects in India? From what I >have seen personally and have heard from others, you do not get a fully >functional application,

In our case, they came close, but it left alot to be desired.

>it doesn't happen in 1/3 the time,

Not even close, it took them 2.5 times as long as their estimate.

>and it isn't even cheaper by the time it's done.

In spite of everything, it probably was cheaper (for us). I can't believe the contractor made anything on this project.

>There are geographic, communication, and cultural factors which few seem to anticipate or adequately manage. Complicated algorithmic logic in non-CRUD applications is particularly problematic. You can just about bet it won't be ported correctly or completely. And that's the heart of your proprietary software.

You hit it on the nose. Their client-facing reps were first-class, but many of their coders were not.

I won't go into all the problems, but 18 months after final delivery we finally have most of the problems resolved. I am convinced that the only way for this kind of project to succeed is to send a programmer/analyst/subject matter expert to the contractor location to assist from that end. The 24 hour work model just did not work well with complicated business rules. They had our source code, but still had many problems.

Our attemp at this was further complicated by going from a fat-client to a web-based app. Our business analysts had a hard time making the switch (why can't we do this in a web app, it works like this in our old app) So the contractor would try and replicate the behavior, thereby killing performance of the web app. I expected more guidance and advice from the contractor, that was one of the reasons we hired them. I could go on, but you get the idea.