[BXXPwg] updated I-Ds submitted

Marshall Rose mrose@dbc.mtview.ca.us
Mon, 11 Sep 2000 23:45:09 -0700


This is a multi-part message in MIME format.

------=_NextPart_000_0065_01C01C4A.51D8E360
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit

i have sent updates of draft-ietf-beep-* to the I-D area. these reflect the
new syntax to support one-to-many exchanges. please review the drafts
carefully and comment accordingly. in the interests of time, i'll post the
unofficial HTML versions to the mailing list in a few minutes. please wait
until the official textual I-Ds are available before sending detailed
comments (so you can include page numbers).


going through my list of issues, the following items remain. i await final
discussion/direction with respect towards making these changes:

1. changing the initialization element from nested XML to character data.

the plus is that this allows non-XML profiles to exchange initialization
data when a channel is started. the minus is that it introduces a second
round of parsing for XML profiles that exchange initialization data.

2. having a per-channel default for Content-Type.

the plus is that this probably saves ~40 octets per message. the minus is
that it makes writing third-person tracing tools harder because to know the
default you need to see the exchange that started the channel.

3. extensible trailer

this was not raised on the mailing list, but was discussed in the hallways
in pittsburgh.

at present,

    trailer = "END" CR LF

it was noted that if the framing format is going to support new features,
then it probably needs some place to put options, parameters, etc. the
logical place is in the trailer. bxxp already has a mechanism for
advertising optional "features" in a greeting.

    trailer = *(directive CR LF) "END" CR LF
    directive = atom *(SP word)
     ;; c.f., RFC822 for the definition of atom and word, except that atom
may not be "END"

basically what this says is that after the payload you'll see 0 or more
directives (each on a line by itself) followed by "END" on a line by itself.
in the case where there are no directives, the trailer looks the same as it
always did.

/mtr


------=_NextPart_000_0065_01C01C4A.51D8E360
Content-Type: text/html;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=3DContent-Type content=3D"text/html; =
charset=3Diso-8859-1">
<META content=3D"MSHTML 5.50.4207.2601" name=3DGENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=3D#ffffff>
<DIV><FONT size=3D2>i have sent updates of draft-ietf-beep-* to the I-D =
area.=20
these reflect the new syntax to support one-to-many exchanges. please =
review the=20
drafts carefully and comment accordingly. in the interests of time, i'll =
post=20
the unofficial HTML versions to the mailing list in a few minutes. =
please wait=20
until the official textual I-Ds are available before sending detailed =
comments=20
(so you can include page numbers).</FONT></DIV>
<DIV><FONT size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT size=3D2>going through my list of issues, the following items =
remain. i=20
await final discussion/direction with respect towards making these=20
changes:</FONT></DIV>
<DIV><FONT size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT size=3D2>1. changing the initialization element from nested =
XML to=20
character data.</FONT></DIV>
<DIV><FONT size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT size=3D2>the plus is that this allows non-XML profiles to =
exchange=20
initialization data when a channel is started. the minus is that it =
introduces a=20
second round of parsing for XML profiles that exchange initialization=20
data.</FONT></DIV>
<DIV><FONT size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT size=3D2>2. having a per-channel default for =
Content-Type.</FONT></DIV>
<DIV><FONT size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT size=3D2>the plus is that this probably saves ~40 octets per =
message.=20
the minus is that it makes writing third-person tracing tools harder =
because to=20
know the default you need to see the exchange that started the=20
channel.</FONT></DIV>
<DIV><FONT size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT size=3D2>3. extensible trailer</FONT></DIV>
<DIV><FONT size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT size=3D2>this was not raised on the mailing list, but was =
discussed in=20
the hallways in pittsburgh.</FONT></DIV>
<DIV><FONT size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT size=3D2>at present,</FONT></DIV>
<DIV><FONT size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT size=3D2>&nbsp;&nbsp;&nbsp; trailer =3D "END" CR =
LF</FONT></DIV>
<DIV><FONT size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT size=3D2>it was noted that if the framing format is going to =
support=20
new features, then it probably needs some place to put options, =
parameters, etc.=20
the logical place is in the trailer. bxxp already has a mechanism for=20
advertising optional "features" in a greeting.</FONT></DIV>
<DIV><FONT size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT size=3D2>&nbsp;&nbsp;&nbsp; trailer =3D *(directive CR LF) =
"END" CR=20
LF</FONT></DIV>
<DIV><FONT size=3D2>&nbsp;&nbsp;&nbsp; directive =3D atom *(SP =
word)</FONT></DIV>
<DIV><FONT size=3D2>&nbsp;&nbsp;&nbsp;&nbsp; ;;&nbsp;c</FONT><FONT =
size=3D2>.f.,=20
RFC822 for the definition of atom and word, except that atom may not be=20
"END"</FONT></DIV>
<DIV><FONT size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT size=3D2>basically what this says is that after the payload =
you'll see=20
0 or more directives (each on a line by itself) followed by "END" on a =
line by=20
itself. in the case where there are no directives, the trailer looks the =
same as=20
it always did.</FONT></DIV>
<DIV><FONT size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT size=3D2>/mtr</FONT></DIV>
<DIV>&nbsp;</DIV></BODY></HTML>

------=_NextPart_000_0065_01C01C4A.51D8E360--