World: r3wp

Kaj yeah but in that time we had some solutions too freebe and r# 
and you own project before boron  ...

You seem to think I did a clone project, but I never did

I mean spinoff existed yet and they runed short because of carl promesses 
can we say at least that this kind of thing is over and that what 
ever happend those actual spin off project will perdurate as real 
intent to break through and break free ?

you didn't do boron and the one before it ? I didn't said clones 
I said spinoff

You seem to be confusing me with Karl Robillard for quite some time

possible

The way I see it, it's fairly simple. We have no control over REBOL. 
The clones are open source, so we do have control over them. If they 
go nowhere, it must be because we're not interested in them, so we 
would only have ourselves to blame, which is much different from 
REBOL

those letters here are so small I can take a j for a l

There's also an extra R, and if you click on me, you can even see 
my surname

Kaj exactly ...  and that's already my case  I will not spend all 
my free time trying to bring up a langage that its author will abandone 
... basically that's why I stoped participating in rebol

That's logical, but open source is different

opensource was the case for the other rebol intent ... as far as 
I know they never succeed in creating a motion around them ... and 
some were really interesting

Yes, so we have ourselves to blame for that

yeah but we don't ...

But if you don't see motion around Red, even more is wrong with your 
eyes ;-)

we do as those never happend and we go on ... same discution over 
and over and over ... even the guy spam his love for jesus growed 
tired of it :)

Kaj I see motion around red  ... and red is really different from 
rebol ...

and I don't understand red ... well I didn't spent too much time 
on it and I don't want to hum be more of  a SoB that I am ... I can 
say I like 3 things in rebol code efficient and short, one small 
VM to do everything, the console to code interactively ... and I 
think this part was underestimated by carl in r3  ...

one of my dream is to have a programing code that allow you to live 
code and see the result of it ... something like the wysiwyg interface 
for the documents formated texts

it's hard to express ... but I really like the possibility to test 
live in the console bunch of codes before integrating in my code 
structures ...

I think over all the language I use or used rebol is the closest 
to this ...

ok so enough of me ranting for this month see you in september if 
I didn't totally forget about rebol til then ( I know you would all 
be over rejoiced by that eventuality :) )

I think rebol in some way could be more efficent with less [  ] ... 
I always found them to much present in my code and I think the python 
way to organise the functions with indentation is interresting not 
perfect but interresting ... not really suitable for the powerness 
of a line of code of rebol ...

hi

A bug?

>> a: [false]
== [false]

>> either a/1 [true][false]
== true

No, try this:

a: [#[false]]
either a/1 [true][false]

I solved it using 
either get a/1 [true][false]

Another way of ensuring that block contains what you think it contains:
   a: reduce [false]

this is typically a source of mistake ...  could be cool to take 
throw somewhere a note on this kind of misleading behavior

Yep, to see the reason:
>> a: [false]
== [false]
>> type? first a
== word!  <<-- it's a word not logic

>> a: reduce [false]
== [false]
>> type? first a
== logic!  << now it's ok

this mean the false isn the block isn't "interpreted" .. yeah the 
reduce need to convert it to what it is supposed to be  makes it 
clear ....

is it just for logic in a block or is it the same for anything ? 
like [1] or ["a"] or [1.2$]

>> type? a/1
== integer!

>> a: ["a"]  
== ["a"]

>> type? a/1 
== string!

>> a: [true] 
== [true]

>> type? a/1 
== word!

hum apparently only logic! are affected

it is like that for the intermadiate values (I don't know if it is 
the correct word in english..)

>> o: context [a: 1]
>> type? first [o]
== word!

cause o/1 =>>> context and context is a word no ?

no, first [o] returns 'o which is a word.

it is different for integer, string, decimal which are not words 
so they don't need to be REDUCED.

It's also same for datatype!. I forgot that words are not reduced 
in block.

Words also represent functions, libraries, and other things. So when 
you put those words in blocks, they're just words:

>> f: does []
>> type? first [f]
== word!


You can say, the words inside blocks are not looked up and not evaluated.

It is probably best to tell, that


[false] corresponds to [one] (provided one has been defined before 
to be one)

, while [#[false]] corresponds to [1]

Similarly, of course, [none!] corresponds to [one], while [#[datatype! 
none!]] corresponds to [1]

Related to the discussion in Core, the rules around making functions 
have changed in R3. Function bodies are not copied any longer:

>> b: []
== []
>> f: make function! reduce [[] b]
>> f
>> insert b 42
== []
>> f
== 42


As I understand it, one reason is to be able to change the functions, 
after they've been made. Why isn't it good enough to manipulate the 
body block before making the function?

Because the body block is affected by the fact, that the function 
was made (binding).

As I side note, I don't fully understand the argument, as you can 
change function in R2 too, after they've made. Do that by change 
the body block, which can be made available using SECOND :function 
(doesn't work in R3 though).

you can change function in R2 too

 - sure, but that is based on a property of another function, which 
 is not desirable

As I see it, the rule should be, that either you wish the function 
you create to be modified/modifiable and then you should use the 
property of the MAKE function, or you don't wish that, and in that 
case other functions should not help to do it.

I understand.

OK what do we have to do to put an end to this masquerade ???