Check out this inconsistency in Lucee where function argument defaults aren't applied when • the UDF was called via

argumentCollection

with not all args supplied • named arguments were used • full null support is enabled https://trycf.com/gist/b5724b0207694e00ed506327a066332d/lucee5?theme=monokai

oooh, scary

Functions in a CFC are a UDF 😁

but when I saw named params and positional params behaved differently

I think this nails it as a bug. I see where you were coming from initially, but something is off if the two ways of specifying the function arguments differ.

It looks like this is caused by the way difference in the way Lucee handles named arguments vs positional arguments. Here's a Gist which adds a dump to the arguments scope: https://trycf.com/gist/9e7095d32923a4c05e0b81b91556bcdd/lucee5?theme=monokai When you run that, you can see that when named arguments are passed, the function dumps: arg1=456 arg2=null But when positional arguments are used, you just get: arg1=256 There is no arg2. This seems like a bug to me. Even when named arguments are supplied, if an argument is undefined then it should stay undefined. It's one of the reasons why the concept of

null

and

undefined

in a scripting language like JS have always made sense to me (and I know that's contrary to what a lot of devs think). I see

null

as a way to say "I'm aware of this variable, but I have no idea what it's value should be" and

undefined

is saying "What? I've never heard of such a thing". In this case,

arg2

isn't

null

it's

undefined

. It was never defined in the arguments and has no default value.

(and I know that's contrary to what a lot of devs think). I see

null

as a way to say "I'm aware of this variable, but I have no idea what it's value should be" and

undefined

is saying "What? I've never heard of such a thing".

I don't know a single dev who doesn't think this. Except some muppets from the dark history of CFML, I mean.

x = javaCast("null", "")

Yeah I think that only got added to CFML in CFMX6

and: function createNull(){ return; }

CF5

CF5 was written in C++. I don't think there was any concept of

null

from that time period.

yeah this is my point. CFML didn't have any need to support a concept of an expression being null prior to CF5. Possibly. And they probably forgot to do anything about it in CF5 and CFMX6. And then... ballsed it up

Thanks for the input, I'll enter a ticket

I put my comment from above into the issue, just so it's preserved