Dioscuri emulation for digital preservation Jeffrey van der

Dioscuri: emulation for digital preservation Jeffrey van der Hoeven Koninklijke Bibliotheek (KB) National Library of the Netherlands we. Preserve Lisbon, Portugal September, 2007 digital preservation research and technology

The cause of all trouble… Refs: http: //www. ecoaction. com. au/res/Image/junkedcomputers. gif http: //www. sinometrecycling. com/Commun/Goods/scrap%20 computer%20 h ard%20 disk. jpg. JPG 2

Digital preservation action! Three basic criteria for defining action: q Policy of the institution q (Future) user requirements q Kind of digital object: Content Behaviour Structure Context Appearance 3

Migration & emulation Migration EXE Pro/ E Original platform Emulator Current platform 4

Emulation = adapt the computer environment to render the digital object authentically. q Pros q Rendering of original computer environment q No changes to digital object q Proven technology q Cons q Complexity q Initial costs (effort) q Isolated process limiting information reuse q Knowledge of original environment required Never applied to an operational digital archiving environment. 5

Emulation Expert Meeting 2006 q Held on 20 October 2006 q Focused on emulation for DP q Attended by selected group of experts in the field of digital preservation, emulation and IT. q States that: q “Emulation is a vital piece of the puzzle for retaining longterm access to the wide range of digital objects. ” q “…important steps to be taken to make emulation appropriate for DP. ” Full statement can be downloaded from www. kb. nl 6

Project outline 2004 KB preliminary study: feasibility emulation 2005 New DP-proof design: modular emulation. KB and Nationaal Archief start joint project. Goal : build and test modular emulator Scope: PDF, databases, multimedia apps. 2006 Tessella leads development. Jeff Rothenberg supports project. 2007 First release of modular emulator. On July 1 st, Dioscuri becomes part of Planets. 7

What is modular emulation? Two key features: durability and modularity q Durable -> emulator has to endure time. This can be done by making the emulator portable to a wide range of computer platforms using a Virtual Machine (VM). Windows 2000 Linux Emulator 1 Emulator 2 Virtual Machine Platform X Platform Y 8

What is modular emulation? q Modular -> emulator consists of modules. Each module emulates the functionality of a hardware component. This way, the modular emulator can be configured much like a real computer. Modular emulator CPU HD RAM Graphics CD-ROM 9

Richard Bill Jeffrey Jeff Hilde Bram Remco 10

11

12

CPU Memory Graphics card PIC RTC BIOS 13

Module. Processor Module Memory Module. Video Module PIC RTC Module Clock CMOS 14

Results q Dioscuri – modular emulator for digital preservation q Current version: 0. 2. 0 (Beta) q Programmed in Java using JVM q Capable of: q Running MS-DOS, Free. DOS, Linux 16 -bit (ELKS) q Norton Commander 3. 0, Word. Perfect 5. 1, Draw. Perfect 1. 1, many games like PC-versions of Pac. Man, Tetris, Chess, Ironman and many more. Even DOS-based webbrowser Arachne works! q XML-based module configuration q Text extraction from emulated environment into the clipboard of host computer. q Running on many platforms like Intel Windows, Power. PC Mac, Sun Sparc Solaris. 15

16

Things to improve q Performance must be increased (limit of 100 x slower) q Data extraction and insertion q More modules: q improved CPU q Mouse q Sound q Network q… q Module library q Replacing JVM by more universal virtual machine 17

Next steps Planets KB 2007 Improve Dioscuri (32 -bit, sound, etc. ) Test elaboration with Dioscuri in reading rooms 2008 Experiments and module library First case: combining emulation with web archiving 2009 Integration with interoperability framework (IF) Integration with e-Depot process flow 2010 Emulation service 18

Future situation Rendering process Digital object + metadata characterisation other service Emulation service Target environment emulator Host environment Emulator and target environment Format registry Software repository Tool registry Contains emulators and environment software Network interface 19

Things to note q Software repository q Disk image preparation q Old documentation (manuals, tutorials, tips & tricks) q Service invocation and execution q Who is responsible for maintaining service? 20

http: //dioscuri. sourceforge. net 21

Questions? References: Dioscuri website: http: //dioscuri. sourceforge. net Emulation project KB-NL / Nationaal Archief and Emulation Expert Meeting (EEM) 2006: http: //www. kb. nl/hrd/dd/dd_projecten/projecten_emulatie-en. html Contact: jeffrey. vanderhoeven@kb. nl 22