During the final year of my undergraduate degree, for my Undergraduate Dissertation, I performed a research study into version controlled content delivery. The results of this study are detailed within my dissertation, but the general gist is that I developed a web service that could host version controlled files of any format and deliver them via REST to any internet enabled location.
The study was primarily focussed on reducing the risk of cyber-attack by automatically closing attack vectors as they were found. What I ended up with was a service that if it had implemented wholesale across the cloud at the time, we would never have heard of Heartbleed as it would have been patched once and as soon as any web page that uses it is refreshed, it’d be updated.
My fringe studies into this new way of delivering content brought up some interesting results. Using the Managed Extensibility Framework, I could host version controlled plugin and extension assemblies. Using Entity Framework, I could import entire Code First data model assemblies. With ASP.NET MVC, using the PreApplicationStartMethod, I could import entire MVC Areas into a site as a single assembly, complete with client-side files such as views, JS and CSS.
À propos of this, it is possible to create extremely thin .NET client applications, given the assumption of a nailed-up connection, that take most, if not all of their functionality from the cloud. Entire miniature websites can be developed as MVC Areas and imported directly into multiple sites within the cloud. Rapid rollout of new content via centralised version controlling. The possibilities are far reaching and I seriously think this needs more study. I shall continue to test different hypotheses but I’d be interested in any feedback on the whole concept of CDNaaS. What difficulties would be faced? How could it be implimented on a global scale? For more background information, please contact me for a copy of my dissertation.