World: r3wp

makes the community aspect of R3 ubiquitous.

for Docbase and curecode it's only open the browser

Yup, that's why it's a link. RebDev is not going to be the final 
name for that though - I think it will be called "DevBase Chat".

lol...
docs: func [
	"Browse on-line documentation."
][
	print "Opening web browser..."
	browse http://www.rebol.net/wiki/R3_Alpha
	exit
]

mwahahaha... very impressive...

Why not? It works :)

yeah it works ;-)

Those docs will be web indexable too, so they'll show up in search 
engines.

And the CureCode access willl let yo see bugs when REBOL itself is 
too buggy to see them. I find that comforting.

I found the new fully integrated REBOL editor.

editor: func [
	"Write your code."
][
	print "Opening Rebol editor..."
	launch "notepad"
	exit
]

Again, whatever gets the release out the door. At least it's mezzanine 
- you can change it in your %user.r.

The goal is to get more developers involved. That is why we focused 
on communications infrastructure: so we can handle more.

don't worry Brian, i think it"s good ideas

even if it makes me smile a little

You might be shocked how much effort it takes to keep REBOL simple 
:)

(btw for the others, the notepad editor is a joke)

Well, you got me - I hadn't used the internal editor in years, so 
I just took your word for it :)

Though I should have realized when you said launch instead of call.

and with no parameters too, y

I'm still waiting on an answer about that INSERT of unset! values 
change in the 2.7 series...

what should be the usage of unset! values in series ?

I don't know. I know that if you need to have an unset! vale in a 
series, not being able to insert one would be a hassle. I just don't 
like the wrapper code  that it requires sometimes (as in that MAP 
discussion).

you can insert unset! as a word , it's enough

Then it is a word, or perhaps a datatype. Sometimes you really need 
an unset! value - like if you *want* an error on eval.

in many scripts i use unset! values to compose optionnals data (like 
in to a draw block).
>> compose [ a b 1 (opt1) 2 c x (opt1)]
with opt as empty blocks ot containing optionnal values

That's nice, I agree :)

so it's a problem

If you compose an unset value, it still doesn't insert it.

ah !?

right !

but not for reduce [()]
it's inconsistent

IMHO

Right, it's supposed to be - that's why there are  two functions.

In 2.7.6:
>> compose [a b c (()) d]
== [a b c d]
>> compose [a b c (#[unset!]) d]
== [a b c d]
>> compose [a b c ([#[unset!]]) d]
== [a b c unset d]

>> compose [a b c (reduce[()]) d]
== [a b c unset! d]

Yeah, I just tested that in R3 too. Note that unset! values mold 
differently in R3. The serialized constructor doesn't work yet either.

Brian, the change in INSERT .. unset! may have been requested by 
Ladislav, but I don't really remember. I suspect it was intentional.

is there a newer meaner version of protect?

one that CANT be unprotected?

cause right now, we can't ensure a word is never replaced by a malicious 
piece of code.

For example, I'd add quite a few protects in my user.r to ensure, 
no application tries to replace how LOAD and SAVE function...

Not yet. Modules are what is going to protect against malicious code, 
in theory.

but what about the actual global words....

I want to be able to protect many core words.

or rather, I'd like for those words to be protected, for my clients 
sake.

for example.

Once modules are implemented there will be no global words.

ok then, I'll keep my question for that phase then  :-)

Please ask them then!

Here's my favorite new REBOL 3 function: WHY?

>> 1 / 0
** Math error: Attempt to divide by zero
** Where: /
** Near: / 0

** Note: use WHY? for more about this error

>> why?
Opening web browser...