red/red

Red Programming Language main room

theangryepicbanana

@ALANVF

this is of course how I designed my language

though I did go off the rails a bit before, I still think it's important to look at all sides of any design issue

Gregg Irwin

@greggirwin

Yes, and you will find that we are generally open to alternate views. Changes do have to be strongly defended though, because so much thought has gone into many aspects already. There are always gaps of course, and things done for expedience or even accidentally.

hiiamboris

@hiiamboris

On style, here's how I write alt-rules https://codeberg.org/hiiamboris/red-common/src/branch/master/composite.red#L60

Gregg Irwin

@greggirwin

It can be hard work, struggling with these choices, but also rewarding and enlightening. If it weren't for all the people I've debated with and learned from, I wouldn't be much of a programmer, or a person.

theangryepicbanana

@ALANVF

@hiiamboris sure, but it still doesn't look as great because | often acts not only as a divider or "indentation", but also as a way to identify the beginning of a branch

yeah I get that

Gregg Irwin

@greggirwin

Ah, composite. When Boris and I are old men we will still debate and change sides on the macro-or-not view. Just when he almost had me, he said "But there is a use case..." ;^)

Gregg Irwin

@greggirwin

Here's a fun thought experiment for you @ALANVF, which you can tinker with in code. Could you add a "begin" operator to do what you want? e.g.

parse "abc" [
    ! "abc"
    | "def"
]

theangryepicbanana

@ALANVF

well I've thought about doing this, something like !: [] would probably work for that just fine

Gregg Irwin

@greggirwin

You're quick. :^)

theangryepicbanana

@ALANVF

I mean it works, but it's a workaround for a simple feature (which is something I highly despise in languages)

I feel like there are better solutions

Gregg Irwin

@greggirwin

It's not a workaround, it's Red being flexible enough to let you do what you want to do, even though it disagrees.

If you feel like there are better solutions, by all means, write them up.

Try not to judge languages too harshly. Workarounds are the compromises of programming, and without some level of compromise it's hard to accomplish anything.

To be clear, you do understand that your view is not "right" here, yes? It's just your view, and a different choice. Not better, just different.

theangryepicbanana

@ALANVF

yeah

Gregg Irwin

@greggirwin

:+1:

theangryepicbanana

@ALANVF

it's just how I see things. I do try not to judge design choices to harshly, but in the case of languages like Haxe, there becomes enough "workarounds" and such that it becomes painful to accomplish anything

Gregg Irwin

@greggirwin

You'll find agreement here that not all languages are equally well designed. ;^)

theangryepicbanana

@ALANVF

of course

Red can get away with it because it's a very powerful homoiconic language, however many languages are not and still try to act like it's no biggie

however it can get to a point where you spend more time working around the language than working with it

GiuseppeChillemi

@GiuseppeChillemi

So, the rule is: if the function is on the stack, its context is available to other functions, otherwise you get a context not available error

f1: func [arg] [f2: func [x][probe (get x) * 2] f2 'arg] 
>> f1 22
44

f1: func [arg] ['arg] 
f2: func [x][probe (get x) * 2]
>> f2 f1 22
*** Script Error: context for arg is not available
*** Where: get
*** Near : x
*** Stack: f2 probe  

>>

Oldes Huhuman

@Oldes

I don't understand, what you want to do, but in your case, first is evaluated f1 22 with result arg, which has no context.

>> f1: func [arg] [:arg]
>> f2: func [x][probe (get x) * 2]
>> f2 f1 22
44
== 44

(at least in R3, does not work in Red, but you probably want something totally different anyway)

GiuseppeChillemi

@GiuseppeChillemi

I simply want to underline when the context of ARG is available, as rule. If the function has not returned, you can pass one of it words to another one, otherwhise, no context is available.

I have written this, so if there is an error in this rule, someone will correct me.

Gregg Irwin

@greggirwin

Not everyone here monitors the red/gui-branch room, but check this out: https://gitter.im/red/red/gui-branch?at=628754784aa6c31dca0f02e4

GiuseppeChillemi

@GiuseppeChillemi

Yes, there are a lot of subtle differences between Red and Rebol.

Red

>> a: ["hello" world]
== ["hello" world]
>> a/"hello"
== world
>>

Rebol

>> a: ["hello" world]
== ["hello" world]
>> a/"hello"
** Syntax Error: Invalid path -- a/
** Near: (line 1) a/"hello"
>>

A lot are really nice!

theangryepicbanana

@ALANVF

wait, that's a thing in red? I thought that was only for hash!s

GiuseppeChillemi

@GiuseppeChillemi

Yes, in Rebol you must use: a/("hello")

In Red you have more freedom

theangryepicbanana

@ALANVF

ah ok

theangryepicbanana

@ALANVF

curious, what's the purpose of the more member for function!s? https://github.com/red/red/blob/master/runtime/datatypes/structures.reds#L263

is it the body of the function?

ohhhh it's a node of other fields

@mosee101

hello, is there a way for Red/Rebol to detect a keypress down on keyboard, and after detecting the keypress, have red press another keyboard key...can i also do this with the mouse clicks?

gltewalt (The other Greg)

@gltewalt:matrix.org

[m]

Yes, maybe, yes

Greg T

@gltewalt

view [base 300x300 "Type some keys, and view your console." on-key-down [print mold event/key]]

Greg T

@gltewalt

view [
    base 300x300 "Type some keys, and view your console." 
    on-key-down [if event/key = #"L" [print "You found the L key!"]]
]

Greg T

@gltewalt

Here are the event names:
https://github.com/red/docs/blob/master/en/view.adoc#event-names

An actor is a function for an event.
https://github.com/red/docs/blob/master/en/vid.adoc#actors

Example: on-key-down

If you mean something like SendKeys, I believe @toomasv had made someting for Windows

Check out the red/help, and red/gui-branch rooms for more help

GiuseppeChillemi

@GiuseppeChillemi

Another interesting discovery today:

Rebol:

a: [b c]
get 'a/b

** Script Error: get expected word argument of type: any-word object none
** Where: halt-view
** Near: get 'a/b

a: make object! [b: 'c]
get 'a/b

** Script Error: get expected word argument of type: any-word object none
** Where: halt-view
** Near: get 'a/b

Red

a: [b c]
get 'a/b
== c

a: make object! [b: 'c]
get 'a/b
== c

So, in Red you can get path while in Rebol you must use do

The only thing I don't like is the error I receive is I try to access a non existent element:

a: [b c]
get 'a/c
*** Script Error: cannot access c in path a/c
*** Where: get
*** Near : 'a/c
*** Stack:  

>>

It is incongruent with paths result on non existent paths, which none. Also forces me to use the slow try method.

hiiamboris

@hiiamboris

red/REP#113

GiuseppeChillemi

@GiuseppeChillemi

We are facing the same problems...

Where communities thrive

People

Repo info

Activity