Saturday, May 28, 2011

Introduction to LightSwitch

LightSwicth is a new development tool by Microsoft, targeting power-users and non-professional developers, willing to write data-oriented business applications.
LightSwitch can bee seen as a kind of "VB / Access for the Cloud".
The tool is quite complete and easy to use by default but it also provides a lot of customization features.
Here is a good introduction video:
http://channel9.msdn.com/Events/TechEd/NorthAmerica/2011/DEV211
and another one showing extension capabilities:
http://channel9.msdn.com/Events/TechEd/NorthAmerica/2011/DEV354

The LightSwitch pages:
http://www.microsoft.com/visualstudio/en-us/lightswitch
http://msdn.microsoft.com/en-us/lightswitch/default (downloads, examples, trainings...)

It is currently in Beta2 stage, but it is stable enough to be seriously evaluated.

Thursday, May 5, 2011

NoSQL and Windows Azure

An excellent paper by Andrew Burst about NoSQL databases and SQL Azure:
http://blogs.msdn.com/b/sqlazure/archive/2011/05/04/10160671.aspx

Tuesday, April 12, 2011

Trinity: HyperGraph database

MS research is working on an HyperGraph database.
Here is the link: http://research.microsoft.com/en-us/projects/trinity/.


Trinity is a graph database and computation platform over distributed memory cloud. As a database, it provides features such as highly concurrent query processing, transaction, consistency control. As a computation platform, it provides synchronous and asynchronous batch-mode computations on large scale graphs. Trinity can be deployed on one machine or hundreds of machines.


My guess is that Semantic Networks will become more and more important in the future. We will go from a Web of Data to a Web of Information, and the semantic aspect is key here.
Needless to say, RDBMS are not well suited for graph management.
Object databases could be more relevant, but there is no doubt that specialized databases like Trinity could bring something.

Wednesday, February 9, 2011

Morphia = MongoDB + POJOs

Interesting reading on IBM DeveloperWorks:
http://www.ibm.com/developerworks/java/library/j-morphia/index.html

Nice to see more and more people now understand the value of directly using business models as the View part of MVC-oriented data access / manipulation patterns.

Please note MongoDB also has a .NET interface with support for C# and F#.
Would be nice to have an EF provider for it.

Monday, February 7, 2011

JEST: JPA + REST

Interesting article from Pinaki Poddar on IBM DeveloperWorks:

JEST: REST on OpenJPA


It is clearly going in the right direction!

A little bit like SDO, but based on REST instead of SOAP.
"Occasionally-connected" data access and manipulation pattern is exactly what the Xcalia technology was already doing, a few years ago, before being discontinued by PROGRESS Software.

It is nice to see the market maturing towards our original vision.

Thursday, January 6, 2011

2011, the year of data virtualization

First of all, let me wish you all the best for 2011, from both personal and professional perspectives.

Who knows what this new year will bring us in terms of technology in general, and in data services / data virtualization as well.

I have just received a webinar invitation from Denodo: they strongly believe 2011 will be the year of data virtualization. For sure, the concept in gaining momentum, but it is too slowly in my humble opinion. It seems they are doing well with analysts, so let's track them.

http://www.denodo.com/en/resources/webcasts/data_virtualization_2011/mailing.html

Once again, Happy new Year!

Thursday, December 9, 2010

Data is the Next Major Layer of the Cloud

http://www.cloudave.com/8607/data-is-the-next-major-layer-of-the-cloud-a-major-victory-for-startups/

I cannot agree more.
But Amazon is not alone, Microsoft Azure and SQL Azure also provides an efficient and well-integrated solution.

Please also note the brand new initiative from SalesForce.com with database.com

Clearly, data is the sweet spot of the Cloud.

Tuesday, November 9, 2010

Nasdaq Data as a Service

http://blog.programmableweb.com/2010/11/08/nasdaq-continues-the-data-as-a-service-trend/

We'll see more and more business data sources exposed as data services.
This is more or less what Microsoft Azure DataMarket tends to generalize.
Today it is mostly limited to read-only kind of access, but I'm confident data manipulation services will follow soon.

Now, the 2 big questions:
  1. How to widely enable all these data services, 
  2. and how to easily combine these external data services with other heterogeneous data models?
This is exactly where Data Services Platforms have a great role to play.

Thursday, October 28, 2010

Big data

Interesting series of videos by Jeff Jonas, "Mister Data" at IBM:
http://techcrunch.com/2010/10/27/big-data/

See also: http://en.wikipedia.org/wiki/Big_data

Two problems:
1. How to deal with huge volume of data? How to exploit it, understand trends, cycles, etc. ?
2. How to bring sense, meaning into that big data?

For #1, see Attivio, ClickFox or Aqumin for new technologies being able to nicely display huge amount of real-time data.

For #2, connect data together, build a context around it. It's all about semantics.

Tuesday, October 26, 2010

jOOQ

Seen this article on TheServerSide: here.
Product page: http://sourceforge.net/apps/trac/jooq

How should we react when reading posts like this in 2010?
It's like being back 15 years earlier.
How is that possible to write so many irrelevant statements in a single post?

The so-called "pragmatic" approach has given birth to so many useless frameworks like this one.

The good news is there is still a need for a comprehensive data manipulation solution.

Monday, October 25, 2010

JBoss Teiid

http://www.jboss.org/teiid
Seems to me this is a basically a kind of repackaging of their good old  metamatrix offer, with all its benefits and limitations.
Anyway, it seems RedHat is going in the right direction.

Any feedback from user would be welcome.

The importance of Data in SOA infrastructure

Interesting thread seen on LinkedIn: here.

Considering data is key when deploying a successfull SOA.
Sometimes services encapsulating data tend to reinforce silos instead or reducing them.
This is a kind of paradox but this is true.

A real data strategy should be at the heart of SOA, before anything else.
As I always said, data has been the failure of most large architectures: CORBA, EJB, etc.

NB: seems to me in the .NET world, data is now considered as a first citizen.

Restart that Blog

I haven't posted anything here since more than one year.
I have been in some personal turmoil after the acquisition of Xcalia by Progress Software.

For different reasons, I think it is the right time to restart communication around Data Services.

I have just created a LinkedIn group about Data Services, feel free to join and contribute.

And stay tuned.