Multi-block message bodies

[Stuart Booth]

On Tue, 10 Jun 2003 17:47:26 +0100, "Kevin Hawkins"
<<a
href="/group/xAP_developer/post?postID=XmfMD5rqDUTAwMssBn8cq850WLUWr7MEWvb3bFNZqlCovI-7Nl-Sl_tHV2_-D48fN0N0ColI5SKxbN8fDwk_">lists@u...</a>>
wrote:

>And re Stuarts comment below - any feedback ? - I personally am sort of
>against having multiple bodies as the way to do this mainly because I
am
>considering the size of the packets but maybe it's correct ??

AIUI James wishes each block within a message body to be uniquelly
named, hence Block.Name.1 and Block.Name.2

As it happens I was reading the spec last night for something or other
and I came across an example which references my.Block.1 etc in the
"Wildcarding of address via Body (Multi-Block messages only)"
section.

Seems like you guys have already been doing this type of thing
already.

My own thought was to keep data out of the block name, but I can see
the usefulness of uniqueness in the names.

For myself, I wasn't too wild on indexed data within each block:

Block.Name
{
Data1=This }
Line1=Wibble }

Data2=That }
Line2=Wobble }
}

I think separating this into multiple blocks makes processing easier,
because the possible data in the block is much reduced, as well as
making the block schema much neater, but at the expense of a very
slight increase in data in a message body (ie the cost of repeating
the block name plus a unique index value, and a little bit).

Block.Name
{
Data=This
Line=Wibble
}
Block.Name
{
Data=That
Line=Wobble
}

I hadn't realised I could do block target addressing, but it's easy
for a receiver app to add their own support for this using my
xAPFramework classes, so I don't feel I need to add any explicit
support for this in the SDK.

Anyway, was there any thinking on block name indexing in previous
discussions? I.e. Block.Name.1

It's only needed where multiple blocks of the same name exists. I'm
wondering about adding some support for this in my framework and was
curious to know if there was any specifics about this format e.g. such
things as "must be sequential only within a single message body".

Block.Name.1
{ }
Block.Name.2
{ }

is okay, but

Block.Name.3
{
}

on its own wouldn't make sense.

Or whether the data could be randomly indexed:

Block.Name.1234
{ }
Block.Name.2468
{ }

This seems to be getting towards including *data* with the block name,
which strikes me as being wrong. A sequential (although perhaps
randomly ordered within a body, but preferably sequentially ordered)
list seems useful and something I can add support for in my xAP
framework code.

Just curious.

S

> > -----Original Message-----
> > From: Stuart Booth [mailto:<a
href="/group/xAP_developer/post?postID=Oq9B88MX1V-x5yE5zEWkp7ddx15UfuDQ-ULfF0S32eQucFP-wkkwAIwfD9tIK4SXOK2Dy8EQSgI5x2C-9Ck">lists@x...</a>]
> > Sent: 01 June 2003 18:22
> > To: <a
href="/group/xAP_developer/post?postID=x1s4uO5FKBEN2UiEO1yf0WUI8V7KkRx9t3AtvTOe3sw2Pc9GHHFuMT6oA3kEXrsq4tbrWBiljv9_CVIECTTIdo1Di_GQKHE">xAP_developer@xxxxxxx</a>
> > Subject: Re: [xAP_developer] Topic 1: Base Level Status Schema

[munch]
> > This reminds me of what I've been wondering about James' News
feed. If
> > you included data in the block name you can include the status of
each
> > port in a separate message block in the message body, and use the
same
> > format for each block.
> >
> > Is there any scope of including instance data in the block name
format
> > e.g. Block.Name:Instance ?
> >
> > I've preferred to have data variables inside the blocks though.
It's
> > data, not a type, after all.

--
Stuart Booth
xAPFramework.net - a reusable xAP framework for .net

<a href="http://www.xapframework.net/";>http://www.xapframework.net/</a>
<a
href="/group/xAP_developer/post?postID=dRhPPoLm5P9VbaGn8XCQVCpefTDaRyE1benxiM2F9LRRFPY1jW1H6llFQTYJ6LWqOVFjDVD3c2-JFKgTMw">stuart@x...</a>