The slides don't match at 30min mark "dogfooding": is when a company uses their own product. This can have advantages like working out bugs internally, before publishing it, or show confidence in the product. MongoDB transitioned from a dogfooded database to a VC backed venture. Some notes: Trend in company: First adapt tech that's open source, then those that are free (as in cost) but close source and finally payed closed source stuff. Michael Stonebraker: the "grandfather" of db implementation, his team developed the architecture of modern sql. Turing Award winner 2014 Stonebraker argued that database vendors these days can only survive by offering their software for free (as in beer), accumulate lots of user this way and finally monetize a small number of these users with e.g. support service schemes or freemium (= alls free, but not this shiny feature). This in effect drives Mongo to make their product popular, and not to design it well. So why isn't the market penalizing MongoDB by going elsewhere? Because the switching costs are high due to data being "inertia". That is the problems of bad design might only appear at scale (so later, and thus sucker in lots lots of users). But by the time you notice the problems you've built your business around the DB, with lots of coupling/lock-ins! It's _NOT_ easily replaceable like a module. Unlike SQL, that has lots of competing alternatives, MongoDB is a "NoSQL" data base, not using the relational db model (or "not only" according to wiki), appearing to be more fit for big data and real-time web applications. The VC-growth models are satisfied due to lock-in caused by data inertia (dependence on the data modell solution of mongodb) Same with free media, where the tendency of customers to only consume free media, drivees publishers to attract lots of eye-balls of which they only monetize a small portion. Now the ones that pay should "ethically" (meaning for the sake of oneself and others) encourage the vendor to improve their quality. But the misalignment/counter-force caused by the vc-growth model is that the publisher needs to attract a large number of consumers first. Generally speaking, encourages models of attracting large numbers of eye-balls but then monetizing a small part of it
The slides don't match at 30min mark
"dogfooding": is when a company uses their own product. This can have advantages like working out bugs internally, before publishing it, or show confidence in the product. MongoDB transitioned from a dogfooded database to a VC backed venture.
Some notes:
Trend in company: First adapt tech that's open source, then those that are free (as in cost) but close source and finally payed closed source stuff.
Michael Stonebraker: the "grandfather" of db implementation, his team developed the architecture of modern sql. Turing Award winner 2014
Stonebraker argued that database vendors these days can only survive by offering their software for free (as in beer), accumulate lots of user this way and finally monetize a small number of these users with e.g. support service schemes or freemium (= alls free, but not this shiny feature). This in effect drives Mongo to make their product popular, and not to design it well.
So why isn't the market penalizing MongoDB by going elsewhere? Because the switching costs are high due to data being "inertia". That is the problems of bad design might only appear at scale (so later, and thus sucker in lots lots of users). But by the time you notice the problems you've built your business around the DB, with lots of coupling/lock-ins! It's _NOT_ easily replaceable like a module. Unlike SQL, that has lots of competing alternatives, MongoDB is a "NoSQL" data base, not using the relational db model (or "not only" according to wiki), appearing to be more fit for big data and real-time web applications.
The VC-growth models are satisfied due to lock-in caused by data inertia (dependence on the data modell solution of mongodb)
Same with free media, where the tendency of customers to only consume free media, drivees publishers to attract lots of eye-balls of which they only monetize a small portion. Now the ones that pay should "ethically" (meaning for the sake of oneself and others) encourage the vendor to improve their quality. But the misalignment/counter-force caused by the vc-growth model is that the publisher needs to attract a large number of consumers first.
Generally speaking, encourages models of attracting large numbers of eye-balls but then monetizing a small part of it
Now I must agree that GPT4 is sentient ( as well)!!