Re: worst case time using hash-tables in Lua

[Francisco Olarte <folarte@...>]

Andrea:

On Wed, May 13, 2020 at 11:25 PM Andrea <andrea.l.vitali@gmail.com> wrote:
>> because your comments about embedded systems where a filename in a
> Actually you are right. The company I work for manufactures microcontrollers and MEMS sensors. I would like to push for Lua to be used in embedded system. (if I were marketing I would say "Lua on the edge, as well as Lua on the cloud").

I do not know what limit you have ( I worked with microcontrollers
very long time ago. Things like 8 levels stack, 128 bytes ram and 4k
rom. Even the powerful things got just an 8048/51 ). This days I would
not consider lua for anything not having a proper 32 CPU with MMU, a
megabyte stack and some available megabytes per state, below this, its
the C-path for me ( actually C++, but without using containers,
iostreams and huge thingies, I use it as just a safer C ).

> 3) Documentation is great (as an example https://stevedonovan.github.io/lua-5.1.4/ and the reference manual and the book from Lua creator which I bought) but in some case is not up to date - as an example see the discussion on how hash table works in current release happened just here few days ago (I implemented the old hash tables, "Brent", just to understand them better); to reconstruct the history is possible but takes time (I have read the all archive since the inception), sometime choices could be better justified (see the recent local-vs-global discussion).

Just be careful on taking the manuals as a normative reference, I used
to do that but RI proved me wrong.

Francisco Olarte.
_______________________________________________
lua-l mailing list -- lua-l@lists.lua.org
To unsubscribe send an email to lua-l-leave@lists.lua.org