The purpose of this workshop is to discuss the full "costs of curation" of HEP data, e.g. that from the LHC, over a period of several decades (say 1, 2, or 3).
To look that far into the future, the past can be used as a "guide" as to the scale of the changes that might occur.
25 years ago, the LEP collider was about to start operation - which was shortly followed by an era of significant change in offline computing that continues to this day.
30 years ago, the Z and W had only recently been observed, Wylbur and MVS were still the main central computing services at CERN and central interactive services (VM/CMS, VAX/VMS) were only just making their debut.
The "programming language of choice" was Fortran - albeit often with extensions at the language level (e.g. VAX Fortran) or through libraries (e.g. Hydra, ZBOOK, ZEBRA etc.)
(The default terminal settings - 24 x 80 - are a legacy from this era.)
Storage capacity was minimal (the VXCERN service, originally based on a VAX 8600 or "Venus", had 6 disks of less than 500MB capacity each: one for each LEP experiment, one for the "system" and one for scratch space).
Today, mobile phones have internal storage of several GB, GHz multi-core processors and are pseudo-disposable devices.
What changes will take place in the future?
This is of course unknown, but we do know that there are clear scientific reasons for keeping the data from today's current experiments - as well as those from the recent past - fully usable until well into the future.
This workshop is about establishing the costs of such an exercise.
Having established these costs, according to a number of possible scenarios (e.g. "best case", "worst case"), this information will then be used for resource and budget planning.
