Re:Does it scale? Who cares (2011)
by hliyan on Sunday April 09, 2017 @12:11AM
(#14070402)
This is probably because implicitly made available decisions from the get go. For example, if you had chosen MongoDB instead of MySQL, the story would be different. Ditto if say, Meteor instead of Java.
[ Parent
| on HN ]
Re:Does it scale? Who cares (2011)
by timewarrior on Sunday April 09, 2017 @12:20AM
(#14070444)
I am guessing you point is that if I had chosen Meteor and MongoDB - we wouldn't have been able to scale.
My initial DB schema was pretty bad. Had to do quite a few DB migrations - which took weeks of work to execute.
I have used MongoDB extensively since, and I am confident that it would have helped us scale to 5M users comfortably. I might have migrated to something else at that point.
There are very few products who reach even 5M users. Which is why developers should focus on launching fast.
Another thing is that best and most successful product have a really simple core product (remember Facebook, Twitter, Instagram etc when they had 5M users). It is not that much work to migrate and rewrite. When products are not getting traction - is when they start getting overcomplicated.
Isn't Meteor a framework? I'm assuming you meant Javascript instead of Meteor
Yes, I should have said JavaEE vs JavaScript