As they keep selling some licenses for companies tied to old software and bunch of stuck-in-time coders they wont fully dissapear, even more if we take in consideration that they dropped any effort on ams in favor of andromo
thingie that never worked and something that someone that does mobile apps like me will never use for sure.
They tried to emulate the ams model on mobile dev but just ate a bug because that crap doesn't have lua or any competent metalanguage, and components made by 5rupee/hr dev like theyre doing are just unusable.
Its a bit of shame because they could EASILY overcome lots of ams's lifetime defects, for example using libuv for async operations, like
luv but some of us remember how slowly they finally got lua 5.1 in ams8 while using and pushing for long ago deprecated version and shitty compatibility layer (lua-compat) lua 5.0.
To be honest, lua itself is in bad shape too, since 5.1, great luajit by mike pall appeared but also gone with him leaving the project (and never finishing a promised fancy GC that looked mighty on paper) since then, seems lua 5.3 has reached luajit performance, but also with the characteristic stability of lua vm, but new api were introduced and some was removed so it breaks compatibility and that like spit in the face for some devs that are switching to js or even py bullshit (quite fun because py suffered of schizo community issues and lots of compatibility breakages too).
Bad decisions and very late reactions killed it, just following the mantra if it works dont touch it, well, not working for todays paradigm shift.
