355 karma · joined December 21, 2020
amount of QA and real-world bash runs, exploratory Go programs, code tinkering, research, manipulation, is outstanding
this poor thing did incredible amount of QA and experimentation work to get it right. and it did! very complex task.
required only nudging in the right high-level direction
sometimes it recommends high-level architectural changes.
sometimes it detects "inconsistent" things that step out of the line of non-trivial patterns. saved me tons of time in K8S and other large config YAMLs and in just imperative code too (structures there repeats often too)
and about correctness of suggestions. they are obvious once you see them. tests pass by definition (if you have any, before Copilot kicks-in). so whatever Copilot suggests likely already passes the tests, yet not ... sit right with it. it's intuition is pretty remarkable.
what we see instead is a myriad of half-backed, useless, un-maintained, poorly-executed, poorly-operated, poorly-designed solutions to non-existent problems, like shooting a gun into the sky and hoping it will land on a target.
lots of new bets are technically impossible. unless you doing something super trivial, you will hit roadblocks that require effort and time (e.g. Apple App Store reviews are notoriously slow and can take a month for a single new app).
oh my god. you have just wrapped standard git CLI. well, this is dissapointing.
holdup... is this just a wrapper around git?
maybe there should not be consent screens popups that break? or any popups? or any consent screens? and why there is consent for something anyways? and maybe there should not be any tracking (or at least not broken one)
https://www.scmp.com/news/people-culture/trending-china/arti...