9 comments

Add a comment…

Sign in

prestine

Your email address

Check!

invalid email

(thinking…)

Reset

or sign in with

• 
• 

UserVoice password

Forgot password?

Create a password

Your name

I agree to the terms of service

Signed in as (Sign out)

Close

Close

Post comment Submitting...

• NSanchez commented  ·  November 10, 2010 8:33 p.m.  ·  Flag as inappropriateFlag as inappropriate  ·  Delete

  I would prefer a more space+speed efficient format (binary or a compact text alternative) to communicate between roles (web/work) and save date in the azure storage. XML is too verbose!

• ChrisLamont commented  ·  September 29, 2010 7:50 a.m.  ·  Flag as inappropriateFlag as inappropriate  ·  Delete

  +3 ... a very reasonable request that shouldn't be too tough to implement

• James Andrews commented  ·  May 20, 2010 4:34 p.m.  ·  Flag as inappropriateFlag as inappropriate  ·  Delete

  Support for Json would be simply grand.

  It looks like many developers end up with several serialization methods for both ATOM and Json, this would definitely make the jump to Azure table storage quicker and easier.

• Rob Gillen commented  ·  January 4, 2010 9:43 a.m.  ·  Flag as inappropriateFlag as inappropriate  ·  Delete

  I agree. In our use cases, both JSON and CSV are very interesting. As Gaurav indicated, we've actually taken to building our own azure-side "proxy" that converts ATOM to a "thinner" format for transport over the wire to help with speed.

• Roger Jennings commented  ·  January 3, 2010 9:34 a.m.  ·  Flag as inappropriateFlag as inappropriate  ·  Delete

  In all table population tests I've run, the Atom overhead greatly exceeds the data payload. I suggest a CSV format because even POX can result in 5X to 20X overhead, depending on the data type.

• Neil Mackenzie commented  ·  January 2, 2010 11:12 a.m.  ·  Flag as inappropriateFlag as inappropriate  ·  Delete

  I think feed-based data-transfer models are intersting but they become a problem when the feed overhead consumes more bandwidth than the data.

• Andy Britcliffe commented  ·  January 2, 2010 10:57 a.m.  ·  Flag as inappropriateFlag as inappropriate  ·  Delete

  Agreed - we've moved to a JSON model for all our data transfer as it's just much more "lightweight" than XML and would be great to be able to push that straight into Azure.

• Gaurav Mantri commented  ·  January 2, 2010 10:17 a.m.  ·  Flag as inappropriateFlag as inappropriate  ·  Delete

  At the very least JSON should be supported. Please see the following blog posts by Rob Gillen:

  http://weblogs.asp.net/rgillen/archive/2009/08/14/atompub-json-azure-and-large-datasets.aspx

  http://weblogs.asp.net/rgillen/archive/2009/08/20/atompub-json-azure-and-large-datasets-part-2.aspx

• Perry commented  ·  January 2, 2010 7:40 a.m.  ·  Flag as inappropriateFlag as inappropriate  ·  Delete

  I did a test with transferring 1000 entities from a table it the payload was 2.5 MB! That's an average of 2.5KB per entity when my average entity data (including PartitionKey and RowKey) is 300 bytes. The overhead is too big.