Netflix Quietly Removes A-Z and Other Sorting Filters
https://www.whats-on-netflix.com/news/netflix-quietly-removes-a-z-and-other-sorting-from-web-ui/Open linkView original on sh.itjust.worksSyndicated from the fediverse. Read and engage on the original instance.
View original on sh.itjust.workshttps://www.whats-on-netflix.com/news/netflix-quietly-removes-a-z-and-other-sorting-from-web-ui/Open linkView original on sh.itjust.works
12 replies
I actually have some experience in this area.
The significant majority of titles start with "A ..." or "The ...". So in any simple A to Z sorted library the T and A are huge. For the library I worked on they were more than half the library combined.
Obviously one solution is to sort those titles as "..., A" or "..., The"; but you have to do that as well, so it massively inflates the index.
This is assuming your titles' metadata is actually correct. It goes pear shaped when a show's seasons have different distributors, or licenses, or anything. All of a sudden they are different shows according to your shit metadata, and now show up several times in the index.
So, after fighting with drones from marketing about these problems and not being able to magically fix both objective and subjective metadata problems with software you end up with a crap A to Z user experience anyway.
Heaven forbid you have titles that start with numbers or symbols.
Billion dollar companies struggling with data sanitization for what is honestly rather trivial data isn't really something I sympathize with
Neither. Among the reasons I'm not in that industry any more.
Put me down as not having a problem with huge T or A.
This comment is exactly what I was fishing for.
I actually said that at work at the time which raise a few eyebrows and garnered a few smirks.
I literally learned how to sort shit like this in my high school programming class, and they can't figure it out? Bullshit
Are you honestly suggesting that one of the most sought after engineering employers in the US is removing a basic UI feature because it's too hard?
Not that it's too hard, that even having billions of dollars doesn't fix shit metadata and they're too lazy to spend some money to fix it.
My employer at the time is dwarfed by Netflix, revenue lead than Netflix would spend on a single title.
Normalising metadata is one of the cases where LLMs are actually good at their job.
That employer would find a way to fuck that up too.
I think Netflix's A-Z filter already filtered on the words behind "A" and "The".
I could be wrong but I remember noticing this in some ordering somewhere.
Yeah, they did it quite well... but I assume they have a new senior vice president of bullshit swinging their dick around.