I just read the (fantastic) article detailing the history of Twitter's infrastructure at scale.
https://news.ycombinator.com/item?id=13441932
And though I appreciated it, I'm left wondering whether they really needed to invest all that money and effort? Or were their engineers unconsciously padding their resumes on someone else's dime? After all its difficult to gain such experience or knowledge otherwise.
Maybe in Twitter's case it made sense to exert that much control over their service. Especially at the time they were founded.
But in today's "cloud first" world, it doesn't seem to make sense for any company to roll their own infrastructure [1]. To me it feels like they may have over-engineered their solution. It would be good to know the business motives behind the decision and for the continued use.
I'm afraid as engineers we're biased to "roll our own" solutions, if not for the experience then out of a hubristic thought that what we make will be better than whats existing.
Do you find engineers and tech companies in general (especially in the Bay Area) tend toward this?
Should we stop it? and if so, how?
[1] I don't mean to focus exclusively on infrastructure or Twitter. I see the same with Uber or any number of tech companies who feel they need to roll their own services like the Big Four.
And though I appreciated it, I'm left wondering whether they really needed to invest all that money and effort? Or were their engineers unconsciously padding their resumes on someone else's dime? After all its difficult to gain such experience or knowledge otherwise.
Maybe in Twitter's case it made sense to exert that much control over their service. Especially at the time they were founded.
But in today's "cloud first" world, it doesn't seem to make sense for any company to roll their own infrastructure [1]. To me it feels like they may have over-engineered their solution. It would be good to know the business motives behind the decision and for the continued use.
I'm afraid as engineers we're biased to "roll our own" solutions, if not for the experience then out of a hubristic thought that what we make will be better than whats existing.
Do you find engineers and tech companies in general (especially in the Bay Area) tend toward this?
Should we stop it? and if so, how?
[1] I don't mean to focus exclusively on infrastructure or Twitter. I see the same with Uber or any number of tech companies who feel they need to roll their own services like the Big Four.