#Event is the new Model
##Abstract
Most of us are accustomed to modeling our application domains by their state at any given point in time. When audit logs, reporting views, data migrations and real-time updates are required, our tendency is to bolt them on to this relational point-in-time model, which can be difficult and costly.
Building on the Event Sourcing pattern, we can instead treat each application as an append-only stream of immutable events. In this style of application, the state of the model at any point in time can be derived by replaying all of the events up to that point in time. This also allows for multiple point-in-time representations ("Models") to exist simultaneously in different places.
The Node.js ecosystem provides a particularly interesting platform for this style of application development. With a shared language between a browser-based client and a node-based server, we can even use the same model logic in both places. I'll walk through experiences and lessons learned over the course of the development of an event-sourced real-time browser application using node, socket.io, spine, and related technologies.
Topics include:
I will approach these topics in a retrospective fashion, to share both successes and failures, in the hopes that they might provide some useful learnings.
Also, the code examples will be demonstrated in coffeescript. Get over it.
##Speaker Bio
I am CTO of Substantial, a 60-person digital product studio based in Seattle and San Francisco.
此处可能存在不合适展示的内容,页面不予展示。您可通过相关编辑功能自查并修改。
如您确认内容无涉及 不当用语 / 纯广告导流 / 暴力 / 低俗色情 / 侵权 / 盗版 / 虚假 / 无价值内容或违法国家有关法律法规的内容,可点击提交进行申诉,我们将尽快为您处理。