,
13 tweets,
2 min read
Read on Twitter
So I've been considering lock-in for a day or so ever since @forrestbrazeal's excellent article came out, and I have some thoughts. forrestbrazeal.com/2019/02/18/clo… (1/13)
Forrest asserts that the real lock-in is IAM. I don't believe that's true, upon further reflection. The real lock-in is AWS's dominance in the market. Hear me out. (2/13)
To wit, if I were building a SaaS company from scratch with a pile of VC money that wasn't mine, I'd have a few options. (3/13)
I could use GCP to leverage its amazing--oh wait nevermind, the thing I wanted got discontinued because Google. I kid except I'm not kidding. I don't trust Google's focus on any given product enough to build a business on it. (4/13)
I could use Azure, which is a surprisingly compelling story. The platform isn't a bad choice at all. Except... (5/13)
...I'm going to use AWS. Of course I'm going to use AWS, because I've been using AWS for a decade. So has most of the cloudy portion of the tech industry--and that's the entire point. (6/13)
I've spoken with multiple people who've forgotten more about AWS than I'll ever know about their work with Azure. The learning curve is measured in months. (7/13)
"How do I X in AWS" is invariably going to find results in search engines; the other cloud providers remain hit or miss here. (8/13)
AWS is far from perfect; some aspects can politely be called "complete crap." But I know its warts, I understand what failure looks like, I know where the sharp edges are, and how to work with them. (9/13)
This is the legacy of letting AWS build the cloud without competition for half a decade. (10/13)
It's funny how many companies build exodus plans for leaving AWS (either for on-prem or another provider) yet fail to account for "3/4 of our engineers will quit within six months of project kickoff." (11/13)
Learning a cloud was freaking *HARD*. Learning a second feels to me like a lot of work for very weak marginal utility. (12/13)
I'm not alone in this perspective. And that's the real lock-in that nobody's talking about. (13/13)
