>>12397054you are taking the bait, OP.
MATLAB sells itself as easy to use, but as soon as you start doing anything non-trivial, the language works against you at every step of the way. the language encourages antipatterns. the libraries are buggy. the documentation is trash. and you pay out the nose for it.
the biggest offender is the requirement that you must vectorize code to make it run in an acceptable amount of time. this is a form of premature optimization, which any experience programmer will tell you is an antipattern. the issue is made worse by the fact that code becomes increasingly harder to reason about and modify when significantly vectorized. constant slicing, reshaping, and squeezing dials all this up to 11.
treating everything as a multidimensional array is a bad idea when data is not homogeneous. it hides implementation details. made worse by the above points.
libraries are buggy and poorly maintained. for example, i worked on a project that was almost sunk by the buggy UDP library; MATLAB was dropping 99.9% of packets on a isolated LAN. while UDP is an unreliable protocol, WireShark made it clear MATLAB was getting the packets and dropping them. MathWorks support said it wasn't their problem because UDP is unreliable, but that is inexcusable. realistically the person that wrote the library left 15 years ago and no one knows how to fix it.
the language has a lot of oddities. for instance, types barely work. try adding a double to an int8 and see what you get. it's fucked up.
i used to have a manifesto cataloging things i had come across with behavior, but i lost it with a computer crash. if i thought hard, i could fill up a dozen posts with flaws and reasons why MATLAB is a trash language.
t. MATLAB expert with 7+ years of experience
PS: simulink is for faggots
