Last night I spoke at the Austin Agile monthly meeting on the topic, broadly put, of how cloud computing starts to effect and benefit Agile software development. While I don’t have a recording (I biffed recording my rehearsal of the presentation, tragically, which I’d usually post here), the slides are above.
Aside from quickly going over cloud computing as it’s understood now, the main points I tried to make were:
- The cloud discussion so far has been mostly an operations story. While public web companies have spoken to development concerns a bit, the discussion of what cloud computing means for iteration-to-iteration development hasn’t really started as hot and heavy as the talk sys admins get into.
- You Agile developer types should get more friendly with operations types.
- Make sure your code works in the “sloppy” nature of cloud computing.
- Figure out if a PaaS, as weird as it may seem (not “owning” the entire stack), is a better way to deliver to your customers.
- Look towards user analytics to track what users are actually doing in your application, and feed that back into product management decisions. More far-looking, see if you can do something with the aggregate of that user tracking. Fellow RedMonk Stephen O’Grady generally calls this “telemetry” and writes frequently about it, as in this piece from late 2008.
The audience was sizable and nicely engaged with questions throughout and afterwards – thanks for having me! (And a special thanks to Israel Gat for boot-strapping the arrangements awhile ago.)
I’ll be speaking along the lines of this topic at the Emerging Technologies for the Enterprise conference this April. I’m curious to hear more about what developers – “Agile” or not – are finding in their cloud adventures. Leave a comment, or otherwise contact me if you’re up for chatting.
Disclosure: several folks mentioned in the presentation are RedMonk clients, see the client list.