2014-10-06 14:11 GMT-03:00 Sanjay-M <s...@planage.com>: > what I was saying was supposing we write a server that alone only accesses > the SQLite database. Users behave like clients accessing the server over > TCP/IP. > and the text i pointed to in my earlier mail detailed how such a scenario > could be implemented. > If someone could write a small sample with say one class then it could be > easier to understand / implement
But that has nothing to do with SQLite nor any other persistence solution. In fact most web applications work that way, where the messaging runs over HTTP (and hence TCP/IP) and the server responds to them. NBSQLite3 is a good solution for self contained applications, where you can distribute everything without requiring the installation/configuration of any third party DB server. If you can't have a web server in your network environment (i.e. intranet/LAN) how will you have any other TCP/IP based server? I mean, if you want to run a Pharo image as your server tier, unless you have certain port/policies restrictions, it will make no difference whether you run an HTTP server or your [put your custom protocol here] server. Regards!
