Why open source requires a bigger investment at AWS than its competitors
In the quest to get Amazon Web Services (AWS) to contribute "more" to open source (a dubious yardstick, as I've written), it's also worth evaluating what "more" means, relatively speaking. The same amount of money and contributions from two companies may require different levels of comparative investment. Let me explain.
SEE: Cloud migration decision tool (Tech Pro Research)
Measuring munificenceAWS announced recently that it was joining the Apache Software Foundation as a platinum sponsor, a $125,000 per year commitment. AWS had already joined the Cloud Native Computing Foundation (CNCF) as a platinum member (which costs $350,000 per year for an existing Linux Foundation member). AWS also supports the Python Software Foundation ($70,000 per year) and other foundations, as well as a host of open source events like ApacheCon.
Most of these are relatively new, and AWS' financial contributions to these foundations and events are far less than its primary cloud competitors, Microsoft and Google, have paid over the years.
SEE: Vendor comparison: Microsoft Azure, Amazon AWS, and Google Cloud (Tech Pro Research)
But then, both Microsoft and Google have been printing cash for a long, long time. Just last year, Google minted $12 billion in net income, and has averaged $16.5 billion in net income over the last several years. Microsoft, for its part, has averaged $18.7 billion in net income over the last four years.
AWS? Well, Amazon actually posted a net loss in 2014, and has averaged $1.5 billion in profit over the last four years. Yes, what profit it has stems from a swelling AWS business but, no, the company is not used to spending money that it doesn't absolutely need to spend. (Remember, this is the company that famously used old doors as desks.)
Amazon runs lean. From conversations with employees over the years, the company sees frugality as a cardinal virtue. As such, it arguably is much harder to squeeze a million dollars in sponsorship fees from AWS than it is from its more cash-rich cousins, Microsoft and Google.
Code as cash?The same is true of code contributions, which is usually where critics fault AWS. Microsoft and Google are the top contributors to open source projects, measured in terms of total employees contributing or lines of code committed. Comparatively, AWS is a pygmy.
Well, yes and no.
Again, perspective matters. AWS employees (like Andrew Lusk) have been contributing to Apache projects since at least 2006, and AWS has been increasingly active in CNCF projects like Kubernetes, too. AWS has also started releasing its own projects like Firecracker, which has ginned up support within the open source community. Even so, AWS continues to contribute comparatively less than its peers, and needs to do better.
Yet it's worth considering how much more is required to get there.
Remember that parsimonious culture that permeates Amazon? It's not just about cash. For anyone who has attended AWS re:Invent, the pace and scale of its product announcements is exhausting. The behind-the-scenes work necessary to engineer all that software arguably leaves little time for tinkering on side projects, open source or otherwise. Couple that with the waste-not-want-not culture, and it's not hard to see why developers might struggle to get overt approval for working on open source projects.
SEE: AWS re:Invent 2018: A guide for tech and business pros (free PDF) (TechRepublic)
Not so at Google, which has long encouraged its employees to spend 20% of their time on side projects of interest to them. Or even of Microsoft, which has a rich history of spending handsomely on R&D and, again, has so much cash sitting in its bank account that it can afford a bit of "leisure time." Both of these companies have very different cultures than AWS.
This is not to suggest that AWS will remain a net consumer of open source forever. (Actually, it does, since "net consumer of open source" describes every company on the planet.) As AWS sees more of its future tied up in sustaining and innovating open source, more engineers will be asked to spend their core engineering time on open source contributions, and more cash will be freed up to support the foundations that help open source projects to flourish.
For now, some patience is required. "More" is a relative term, and much harder to come by at AWS than at Google or Microsoft. Even so, "more" is needed, and "more" is what AWS will give as it sees its self-interest bound to open source communities.