top of page

Some smaller parts of DMI still went ahead ...

  • Writer: primalmotion studio
    primalmotion studio
  • Jan 1, 2011
  • 2 min read

Updated: Jan 14

ree

I worked several years as contractor for the BBC: Fabric - Production Tools and Music Reporting System.


Production Tools in Fabric was a large project across multiple areas in-house. This new BBC Production media workflow were aiming to enable Digital Archive and the Media Product Enterprise Systems (i.e.: Sport, Journalism, TV Shows, Films…) integrating A/V Ingesting, Logging and Editing tools. Production Tools project includes: a Distributed Job Processing system, a Cloud Media Storage and EBP Interfacing.


So far so good, the project resolved a number interesting technically challenges in production pipeline at large scale, there were in Production Tools dept some very good skilled contractors and indeed some part of the macro project was working pretty nice and I would say at a competitive commercial level as well.


The bigger the project the bigger the problems. If the reader is coming from the sector, you’ll probably have read how this end up. So, although I’d love to share about some very specific interesting technical problems solved there and learned lessons, it is better to move on into new topics.


However…. :)


… I had the opportunity to participate into another small in-house project (‘small’ considering BBC project sizes): Music Reporting System (MRS). MRS is receiving all TXs from digital and online publications from all UK and Worldwide BBC Radio, TV and Online publications to find matching against the millions of Music & Author records. To put this in simple terms, every time you hear a 'Happy Birthday’ specific version used as as background music in a trailer (for example), by legal requirement, this must be matched and reported automatically.


Early Java Spring years with no NoSQL (because it was not production ready), continuous stream of data coming from TX and complex business logic and semantic matching process against a very large database to find the best candidates. The project migrated and transitioned from previous 10 year legacy system to the new one without any major problem.


Nice hit for the small team there: Some smaller parts of DMI still went ahead, including a replacement music reporting system.

 
 
 

Comments


bottom of page