Profile cover photo
Profile photo
Yrjänä Rankka
Grand Praetor of Excruciations at ZONK.NET Propaganda HQ
Grand Praetor of Excruciations at ZONK.NET Propaganda HQ

Communities and Collections
View all

Post has shared content
Add a comment...

Post has shared content

Post has shared content
Yeah. Apple really dropped the ball on that one.
Add a comment...

Post has shared content
Why an HTML5 based PivotViewer is a big deal

For reasons I'll never understand, Microsoft opted to release a game-changing business intelligence front-end as a Silverlight plugin. Even worse, they opted to keep its future a mystery.

Thanks to +Roger Noble we now have the makings of an HTML5 rendition of PivotViewer, also available in Open Source form, thereby enabling interested parties to continue product advancement etc..

Anyway, why is this a big deal? First off, you couldn't use the Silverlight edition of PivotViewer on an iOS device (iPhone or iPad). Now you can!

I've attached some screenshots that demonstrate the effect of this effort which is highly commendable.

Note: we've done a straight swap & replace of our earlier Virtuoso hosted PivotViewer which adds SPARQL protocol URLs as sophisticated CXML data source names (DSNs), and it works!

#LinkedData #Web30 #BI #PivotViewer #SemanticWeb #BigData #iPad #iPhone #iOS5
3 Photos - View album
Add a comment...

Post has shared content
How programmers see each other... as seen in the AKSW toilets :-)
Add a comment...

Pffft. Software upgrades are overrated anyway. Especially when RH decided to kill a perfectly good virtualization system, Just lost a good deal of hours trying to migrate my Xen guests to KVM after loosing a boot disk on the Dom0 and running into all kinds of esoteric hardware incompatibility issues - it's somewhat less than great motherboard/BIOS.
Ended up reinstalling CentOS 5.8 instead. Try again in another few years of actual productive work. Boy am I glad I'm not using Fedora.
Add a comment...

Post has shared content
Is Open Database Connectivity (ODBC) Dead?

Of course not!

In question though, is the issue of primacy on the value pyramid [1] which fundamentally applies to all SQL RDBMS-oriented technology.

I recently stumbled across an interesting video [2] from Progress DataDirect, our long-time competitor (or co-opetitor, in this Internet induced era of coopetition [3]) that seeks to shed light on the current state of affairs in the ODBC market place. For purposes of this note, I've taken some time-sliced URLs that highlight some key points in this nice presentation.

Q: Is ODBC a commodity technology? Does it have a place in the modern day enterprise?

A: See 7:29, <Is ODBC Driver Technology Dead?>, which highlights the data heterogeneity issue that arises from mergers and acquisitions.

Q: How does the BigData meme affect ODBC?

A: See 9:21, <Is ODBC Driver Technology Dead?>, which hones into the fact that enterprise agility requires insights extracted from increasingly disparate and heterogeneous data sources.

In addition to the response seen in the clip, I would say that the InterWeb inflection has exponentially increased data volume, velocity, heterogeneity, and dispersion. This makes data virtualization critical, and that virtualizatoin can be delivered via ODBC, and much more powerfully via Linked Data.

What's Next?

Imagine an ODBC Driver that lets you:

connect any ODBC-compliant application to a single data source name that exposes tables from several DBMS (Oracle, SQL Server, DB2, MySQL, Sybase, Informix, Ingres, Progress Open/Edge, PostgreSQL, etc.), as though they are all part of a single RDBMS; or

connect any ODBC-compliant application to a single data source name that adds any Web Resource to the DBMS list above -- i.e., not only can you join data across relational tables, but you can do that with Web data (e.g., data from DBpedia) added to the mix!

Just as Linked Data takes you beyond the database-limited connectivity of ODBC [4], imagine being able to use any application that can process an HTTP URL (a hyperlink) to:

use the URL of a Web Resource as your application's Data Source Name -- no ODBC Driver installation of configuration required;

express your data access needs via basic HTTP URLs or more complex SPARQL-protocol URLs -- natively integrating SQL-like query patterns;

share query results, query definitions, views, etc., via hyperlinks -- with no requirement for participants in the value chain to have the same applications installed, and zero additional administrator time;

build your own data junction box in the cloud, that acts as a proxy to all the heterogeneous data sources you can imagine.

All of this may be achieved via Virtuoso [5], which is a full blown CoRelational (Tables and Graphs) DBMS, or simply by using the Virtuoso ODBC driver [6] against public services such as URIBurner [7].

How Does URIBurner & the Virtuoso Sponger aid Open Data Connectivity?

URIBurner is a public instance of the Virtuoso Sponger [8] Middleware. Basically, you use this middleware to turn basic Web Resources into high fidelity Linked Data resources thereby making them SPARQL query friendly. Basically, G+, LinkedIn, Facebook, Twitter, and many others becomes massive tables in a distributed database. You can share your query results, query definitions, and derived reports via hyperlinks.

As super keys, every link is the commencement point for drill-downs using the follow-your-nose pattern i.e., click your way to lots of serendipitous insights culled from a powerful data mesh.


ODBC is no longer the apex of the data access pyramid. Linked Data is a superior solution, for the following reasons:

1. The HTTP protocol separates data from data representation. Any third-party ODBC driver developer would have at some point encountered the frustration associated with not being able to implement wire-protocol drivers; this scenario arises because RDBMS vendors blend their ODBC drivers with a backend network protocol which isn't available to any third party driver developer.

2. The HTTP protocol leverages URI abstraction. As a result, you can integrate a query language into hyperlinks (URLs) thereby turning every URL into a highly potent equivalent of an ODBC Data Source Name, modulo the platform-specific (operating system, wire protocols, data representation formats) shortcomings of ODBC

3. URIs also act as Super Keys. Super Keys go way beyond the limited capabilities of Primary and Foreign Keys, into a magical realm where the Web becomes a DBMS wherein every key resolves to meaningful data that's implicitly easy to reshape and reshare.

Prior to the Web's explosion, ODBC was your best bet. Today, we have Linked Data, which now sits at the apex of the data access, data integration, and even data management pyramid, thereby downgrading (but not eliminating!) the relevance of ODBC.


1. "RDBMS value primacy downgrade is nigh!" -- <>

2. "Is Open Database Connectivity (ODBC) Dead?" video -- <Is ODBC Driver Technology Dead?>

3. "Coopetition" -- <>

4. How Linked Data takes you beyond the database connectivity limitations of ODBC -- <>

5. OpenLink Virtuoso Home Page -- <>

6. Virtuoso Client Connectivity -- ODBC, JDBC, ADO.NET, OLE-DB drivers are standard equipment. <>

7. URIBurner -- Just click on a button of interest! <>

8. Virtuoso Linked Data Middleware ("Sponger") -- .


1. Getting Data into Google Spreadsheets via SPARQL protocol URLs -- <>

2. Getting Data into Excel Spreadsheet via SPARQL protocol URLs -- <>

3. LOD Cloud -- a massive collection of data sources that are available to any Virtuoso ODBC Driver user -- <>

4. Understanding OpenLink Software's Evolution from Open Database Connectivity to Open Data Access & Connectivity -- <>

5. Other G+ Posts about ODBC and Linked Data -- <>

6. Blog Posts about ODBC from the past -- <>

#LinkedData #ODBC #JDBC #RDBMS #DBMS #CoSQL #BigData #BroadData #SmartData #SemanticWeb #Web30 #Spreadsheet #Database #BI
4 Photos - View album
Add a comment...

Post has attachment
Add a comment...

In a teutonic JavaScript hell in Berlin. Not any worse/better than being there in Amsterdam.
Add a comment...

So will @Yahoo be the next one to be utterly SCOrned? Raiding more profitable than R&D? #SwPatents
Add a comment...
Wait while more posts are being loaded