,
10 tweets,
3 min read
Read on Twitter
Gosh, back in 2012/13 when I was saying "OpenStack is a dead duck", I use to get all sorts of grief because a common idea was it would "take over the cloud market" - searchcloudcomputing.techtarget.com/feature/Open-s… ... nowadays, people still claim it's a "success" because of its use in niches.
... let us be clear, OpenStack pursued its "differentiate on API" play leading to a collective prisoner dilemma. The pursuit of differentiation on the one thing that absolutely did not matter - the API - rather than co-opting the existing standard is what led to its problems.
... as I said before, as I still say now, its only path to any future success relies on China and the long game. Even then, it'll just be an invisible subsystem.
Oh, and before someone starts claiming the huge amounts of money spent on OpenStack as a sign of success ... society spends a fortune on Myers Briggs Testing, Astrology, Homeopathy and New Age Crystals. Just because someone spends money on it, doesn't make it a good idea.
There are, of course, a few companies who craftily used OpenStack to their advantage, to extend their existing product lines or to capture a space ... VMWare comes to mind.
X : What about Eucalyptus?
Me : First they weren't OpenStack and second they were acquired. Do you want me to really start listing the OpenStack companies that would take on the world that no longer exist or have re-invented themselves?
Me : First they weren't OpenStack and second they were acquired. Do you want me to really start listing the OpenStack companies that would take on the world that no longer exist or have re-invented themselves?
X : Customers didn't want API compatibility?
Me : Surveys of people bought into the OpenStack idea, the constant messaging that AWS compatibility was a second class citizen / would be dropped combined with an endless narrative against "lock-in to AWS" - let's not talk context.
Me : Surveys of people bought into the OpenStack idea, the constant messaging that AWS compatibility was a second class citizen / would be dropped combined with an endless narrative against "lock-in to AWS" - let's not talk context.
X : I don't see why a focus on AWS APIs would have helped?
Me : You've just answered why ... "a focus". OpenStack needed to focus on what customers were using and more importantly it needed to focus - creating a market of AWS clones could have given it that.
Me : You've just answered why ... "a focus". OpenStack needed to focus on what customers were using and more importantly it needed to focus - creating a market of AWS clones could have given it that.
My last comment on this, is not even my own ... it's from the legendary @b6n (the original source of the idea which became EC2) - "All successful standards are defacto standards" - from OSCON cloud summit, 2010.
... alas the same mistakes are being made in Serverless today. Whilst others are pontificating on building future markets on some standard for underlying components, Amazon just captures actual use and defines the defacto. C'est la vie. That battle is almost over already.
