Objective
This talk will examine the core tenets of what it takes to make your products/platforms truly customizable by letting your customers consume the services offered by your platforms. It offers insights into creating "app"etizers for product platforms using Python to help you along the way.
Description
Let's face it - your products cannot solve every one of your customer's problem. At least not in the state it was shipped out of the factory. Traditional ways of making products field customizable can only go up to a certain point in helping tweak your product to suit the customer needs.
By creating a product platform where people can "run" their ideas, you can not only offer true customization but also let people innovate on top your platforms.
This talk will offer insights into what it takes to serve your platforms using python. It draws ideas from various existing technologies and PaaS providers. In each of the below areas, I will analyze various approaches, available tools/technologies/python recipes, identify best practices etc.,
- Opening up your products * Creating SDK, Abstracting your product's services, APIs
- Creating a runtime * How to service 3rd party code?.
- Process isolation * Ensuring that 3rd party code behaves well. * Virtualizing resources
- Nitty-grittys of running Python code * Environment variables, 3rd party libraries, Native code
- Monitoring your platform * Available tools. Best practices etc.,
- Hardening and Robustness * How to ensure that at the end of the day, your product doesn't crash?
- App stores
If time permits, we will see a small demo (< 10 mins) :
* Introduce a hypothetical product using Raspberry Pi / BeagleBoard
* Create a "serve"able platform (done beforehand)
* Pick an idea from the audience and run it on the platform
Requirements
Basic understanding of Python
Speaker bio
Harish has diverse experience in building systems and applications on various platforms - Mainframes, PC, Embedded Systems and Web. He is been charming "Python" ever since he fell in love with the language about 6 years back.
1
▼
This sounds too wide for a 3 hour workshop. It's necessary to trim it to be concise if the workshop has to be effective and not rambling.
1
▼
Hi
This is not proposed for a workshop. The proposal is for a talk session. My opinion is that I will be able to cover all the important and interesting points withing the time available for a talk and possibly squeeze in a demo too.