Tape is one of these things that is always dead
and always will be nearly dead, according to the 'trendsetters' anyway. Like COBOL, the PC, desktop applications and, what is it this week .... Java, I believe.
Tape library vendor SpectraLogic says it installed 550PB of tape library capacity in the second half of 2012 and reports that its revenues, led by rising T-Finity library sales, for that six months were up 9 per cent compared to a year ago. Half an exabyte of tape equates to the installation of roughly a dozen of the vendor's …
"COBOL is alive and well in financial and legacy engineering firms. If developers want to make scads of cash quickly 'dead' languages are a good place to be."
True, there is a ton of COBOL out there which needs to extended and updated. Unlikely to be going anywhere, anytime in the near future. As no one is trained in COBOL in university (although some electives are now coming back) and the related OSs, utilities, there is a high amount of demand for those skills.
Agree, the fickle IT world seems to think anything which not growing rapidly = dead. Until someone comes up with another way to store PBs of data with very little energy utilization for a very low cost, tape will not even be challenged as the archive medium. EMC was pushing the "tape is dead" idea. Just a coincidence that they don't have anything other than rebrand in the tape market, but are the largest player in the disk and VTL market which would benefit from the decline of tape. Cui bono.