Welcome to how ALL software works. Just about any bozo can crank out crap code. Production ready and scalable? Completely different beast.https://twitter.com/EmilyGorcenski/status/880798373221208068 …
-
-
Replying to @antinazipatrol
Nah, it's more complex than regular software because of living data needs and requirements for low-level maths.
1 reply 0 retweets 0 likes -
Replying to @EmilyGorcenski @antinazipatrol
Perfect example is how in software you can live with versioning and package management. For ds you need to do that for training data, too
1 reply 0 retweets 0 likes -
Replying to @EmilyGorcenski @antinazipatrol
It's more akin to servicing two concurrent software versions that don't work with each other but depend on each other.
1 reply 0 retweets 1 like -
Replying to @EmilyGorcenski
Again, that's a problem in EVERYTHING software related. It's why deployment has become its own discipline these days.
2 replies 0 retweets 1 like -
Replying to @antinazipatrol
Yeah, I'm aware. Bad analogy but it's hard to communicate just how rough shape DS is in. And I've worked in both domains.
2 replies 0 retweets 1 like -
Replying to @EmilyGorcenski @antinazipatrol
So when I say it's an order of magnitude worse, consider that it's literally my entire career expertise speaking.
1 reply 0 retweets 3 likes -
Replying to @EmilyGorcenski
I concur. I also think that even in "regular" software massively short shrift is given to data and data migration/transformation as well.
1 reply 0 retweets 4 likes -
What's the saying? There are only two types of Postgres implementations: those being used as message queues, and those that haven't been yet
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
