I just emplemented word merge in our browser based case management system. I created a directory on each of my user's C drives, c:\webmerge, to hold the MSWord merge documents and the source data text files that the merge documents will use as a data source. My merge documents are programmed to look for the text file in the c:\webmerge folder, and I also require that a copy of the master merge document be manually copied down to the c:\webmerge folder whenever the master template document is updated (which may be every few months or so).
This is used in my application by the user requesting a form from a list on a certain customer's account, and then the Word Document will seemlessly open with the customer's data merged into the document. Behind the scenes, my web server queries the database and creates the souce data text file, c:\webmerge\form119.txt, and then builds the html that will open the specific document, c:\webmerge\form119.doc, and sends this down to the client.
This works fine, however, I would like to move away from my c:\webmerge folder method to possibly a more conventional method of storing these files on the client so that I don't have to create a specifically named folder on each client and I don't have to manually copy the template down each time we modify them... or maybe there's some other way of accomplishing this all together. I definately want to stay away from server-side processing.