Personally, I think there's a lot of FUD in this space.  There are  
cases where it can be an issue, more so from a parsing standpoint  
than a bandwidth standpoint.  Many routers have built in compression  
capabilities these days, and XML does compress very well.  So, I  
don't think the sky is falling in the way that some have portrayed it.

The processing of it is a different concern.  You simply can't take a  
100MB XML file and run it through your favorite parser the way you  
would with a 1KB XML file.  The intermediate structures that are  
created in memory can create significant problems.  Of course, one  
would wonder why a developer would be turning a 100MB file of  
anything into some in-memory object-graph.  It would seem that  
scenario is much more like a batch file processing scenario where the  
XML would be accessed in a random-access fashion, only loading what's  
needed for immediate processing into memory.

So, I think it's definitely something that we have to live with.  I  
don't believe that it is a significant impediment, but I do think you  
need to give some thought to the proper way to process it when you're  
dealing with large documents.  Frankly, there's probably more of a  
risk to corporate network bandwidth due to people watching YouTube  
when they should be working than XML messaging.

-tb

On Mar 16, 2007, at 4:58 AM, Gervas Douglas wrote:

> The question of XML's verbosity has been raised before in this Group.
>  Do you consider it:
>
> (1)  A significant impediment
>
> (2)  Somthing we just have to live with
>
> (3)  (2) does not really matter given the computing horsepower and
> network bandwidth commonly available?
>
> Gervas
>
> --- In [email protected], "Gervas
> Douglas" <[EMAIL PROTECTED]> wrote:
>>
>> <<First we had service-oriented architecture (the familiar SOA), now
>> we're getting "service-oriented infrastructure (SOI)," meaning people
>> in the IT world are starting to think seriously about what it will
>> take to keep SOAs up and running.
>>
>> One approach is grid computing which, at least in theory, can assign
>> appropriate resources to services as needed. Another, focused on a
>> higher level, is the single-vendor SOA suite, where the "plumbing" is
>> left to a single vendor, and heterogeneous interoperability is only a
>> concern at the level of services and applications.
>>
>> One of the biggest infrastructure concerns is SOA's use of XML, which
>> creates problems for networks because it is very verbose. This  
>> creates
>> higher network payloads and also eats up CPU resources because it
>> requires more processing. The problems with XML are so serious that
>> some developers have proposed an alternative, Javascript Object
>> Notation (JSON).
>>
>> But will these problems slow down the SOA steam roller? Few
>> consultants or analysts think so. Freeform Dynamics, for example,
>> recently conducted a survey on SOA take-up and concluded that SOA is
>> "completely unstoppable.">>
>>
>> You can read Stevens's infrastructural concerns at:
>>
>>
> http://www.itbusinessedge.com/blogs/mia/index.php/54/soas- 
> infrastructure-issues/
>>
>> Gervas
>>
>
>
>
>
> ------------------------ Yahoo! Groups Sponsor -------------------- 
> ~-->
> Yahoo! Groups gets a make over. See the new email design.
> http://us.click.yahoo.com/hOt0.A/lOaOAA/yQLSAA/NhFolB/TM
> -------------------------------------------------------------------- 
> ~->
>
>
> Yahoo! Groups Links
>
>
> [EMAIL PROTECTED]
>

Reply via email to