World: r3wp

First quote from section 3.6.3. "Destination address fields"
Second quote from section 5. "Security"

It states in the site that the RFC has been updated to handle security 
considterations... so our email implementation seems to be based 
on an older, out-of-date RFC.

As Gabriele says, 'send doesn't use the header for addressing.  It 
is up to you to implement how email is sent, and to whom.

Sorry, I don't understand those arguements.  From my perspective 
Rebol is sending all the BCC'd recipients when an email is sent. 
 Unless you are telling me it is up to the mail server to not pass 
the BCC along.  I somehow don't think so.

It is the responsibility of the mail client.  You control the mail 
client.  If you construct a BCC field in your header, Rebol will 
ignore that when using send.

Iikewise, for cc: header.

>> help send
USAGE:

    SEND address message /only /header header-obj /attach files /subject 
    subj /show

DESCRIPTION:
     Send a message to an address (or block of addresses)
     SEND is a function value.

It is up to you to determine what the 'address is .. whether it is 
a single email address, or a block of email addresses.

and when you do send to a bock of addresses they are send as individual 
messages (effecively a BCC)

which means the server cannot take advantage of the fact the body 
is the same on all the messages and just spam/virus filter it once.

(having this issue on a  list of 2500)

If you want, you can write your own send command that does what you 
want.

Let's make this clear: SEND is not an email client. SEND is a function 
to send email. If it was an email client, it owuld just show a window 
with the ability to send TO, CC, BCC and so on, with a send button 
in it...

what send does is: taking a message, and sending it to the list of 
recipients you provide. That's what any MTA does too.

the user interface for mailers, though, of course do more than that. 
TO, CC and BCC are part of the user interface.

Hmm.  SMTP allows the sender to make multiple RCPT TO commands before 
sending the body of the message once - which then gets sent to the 
addressees of the RCPT TO.

This is how your email client would implement a CC or BCC.

yep, just like SEND.

I don't know how 'send works ... but I assume it does only a single 
RCPT TO ?

depends on refinements. i don't recall if it does one with a list 
of emails, or several with one email address each.

I think it sends it multiple times.

SEND with multiple recipients sends a copy to each recipient; SEND/ONLY 
sends the same message to many recipients

I don't see the use of using /only, really...  With /show, I see 
all recipients whether or not /only is there - and without /show, 
with /only, it won't show any email address, and without /only, just 
it only shows the recipient.

(without /only, without show for that last one)

(/show)

Quick question .. how do you set the century to-date defaults to 
?

>> to-date "1/10/40"
== 1-Oct-2040

is not what I want.

>> to-date "1/10/1040"
== 1-Oct-1040

but what I want is last century !

Many languages have a way to set the window in which dates work.

WOuld own handlling work? Where do you get the dates from? always 
"1/10/40"?

(i guess there is nothing inbuild, but would be pleasanly surprised 
if i am wrong :)

I want to automatically set a window so I can deal with dates - for 
birthdates and appts.

If I have to, I will have to write my own mezzaine - just was hoping 
that Rebol had something built in.

I guess that step must be made in the conversion, to check if date 
is 2 or 4 digits.

I guess when it was coded, it was set to make everything the current 
century .. after all Rebol was invented last century.

if you know its 2, you can then make a quick check. if date > 2035 
[date: date - 100]

yeah ..

so we need a to-date/window ? How to deal with different formats?

what different formats?

thought to-date understands ultiple formats. if its always */year 
its a short check.

yes, I thought using /year is what is needed.

there are some more mentioned in the word-browser. if always "/", 
if 2 = length? find/last "/" ..

to-date: func [ value /local tmp ][ tmp: to date! :value if tmp/year 
> ( now/year + 10 ) [ tmp/year: tmp/year - 100 ] tmp]

this ensures that years more than 10 ahead get changed

explicit 4 digits are not necessary?

but quick solution :)

Since we don't have control over the field .. users will enter 2 
digits sometimes

i mean user can enter 2025 explicitely. that gets 1925.

well, I see ... he made a mistake.

In my application .. you are not allowed to deal with dates too distant.