The "I Love Lua, but..." post

[thelinx]

Code: Select all

[list]
[*]item
[*]item
[*]item
[/list]

[vrld]

*no control over for loops, sometimes i want a different condition than <= or a more complex increment than addition

Uhm.. what? Generic for.

*arrays start at 1, WTF!! they should start at 0!!

Says who?

[bartbes]

Regarding that little script I posted earlier in this thread, it now has a github repo.

[Jasoco]

Not everyone uses an editor that has text replacing you know. Not everyone WANTS to switch editors from something they like to something that isn't good enough just for that feature. Give me my gosh darn ++ and += and != consarnit!

No, not really a big deal. I've become used to typing it all out.

[slime]

Jasoco, use bartbes' script...

[Xgoff]

The main problem now I guess, is that they have to maintain backwards compatibility with grammar.

No, it's that it would allow ambiguous syntax. In which case it's not at all obvious what the compiler should do. These situations cause debugging to be very hard, because the code doesn't look wrong, and there is probably not even an error at that exact location.

it turns out that lua already has an ambiguous syntax case that's related to that if example:

Code: Select all

local x = y
(z or print)("hello")

it's better not having more of them! (although this case is technically fixed in 5.2, since it'll just assume it's all one statement)

also, if x++ were added, then people would expect x-- to be added too. but... there's a problem with the latter one.

[bartbes]

How is that ambiguous, once the line ends the statement ends, right?

[thelinx]

How is that ambiguous, once the line ends the statement ends, right?

No. Take this, for example:

Code: Select all

foobar = table.concat
{
  "foo",
  "bar"
} -- works!

[Xgoff]

How is that ambiguous, once the line ends the statement ends, right?

line breaks aren't considered significant in lua's syntax

[bartbes]

Right.