Microsoft annouced the new Azure Bot Service on November 15th. The announcement and blog post include the most important highlights.
Now you might think, why don’t I just go to the other blog post? Fair point, but if you want to know what the differences between the old (bad) way and the new (good) way to build bots is, just continue reading. You don’t even have to click on anything.
The architecture needed to build a bot is not completely new, but improved. Especially the new Proactive bot allows to trigger more complex calculations through a build in queue (see image below)
The new approach to building bots comes along with a lot of benefits for developers and IT professionals alike.
No web.config anymore, you don’t even have to press publish. The Azure functions allow for instant changes in you bot environment.
The included chat window allows for instant testing and feedback. When you stop writing code the function is saved and the whole bot can be tested instantly.
Besides working in the browser, which can be really annoying sometimes, the option is offered to use continuous integration to maintain your app code. This option includes local Git repositories, GitHub repositories, VS Team Services and even Dropbox.
Further Modularity and Scaling
Some people have clearly thought that the bots somehow where not modular enough, so they added Azure functions. This allows for a really modular architecture. If you want, you can even deploy other services, like databases automatically with the new bot templates.
Additionally, because we are in the Cloud, we can now scale every service depending on the demand of the bot.