1. Introduction This data set has been generated by the Jet Propulsion Laboratory's (JPL) Science Digital Data Preservation Task (SDDPT) by migrating valuable space science data from older, deteriorating magnetic tapes to CD-ROM. The primary goal of the SDDPT is to reduce the amount of data stored on magnetic tapes, preserve valuable science data and make it more accessible to the user. This data set resides on a series of CD-ROM volumes and contains several types of special Viking Lander archival products generated by the Image Processing Laboratory (IPL). The input for the compact discs was magnetic tape that, in most cases, was at least 15-20 years old. Accompanying each tape on the disc, there is a log file which documents SDDPT processing history. Therefore, if problems were encountered while recovering the data from tape, they will be documented in the log file. The SDDPT is not reformatting the data. However, it has been made readable in standard ISO 9660 format on the compact disc useable on a variety of platforms. Three copies of each title will be generated: one intended for deep archive at the National Space Science Data Center in Goddard, MD; one as a backup copy at the JPL Archives; and one to reside with the data user. In the cases where a planetary data set is likely to have many users, the "user" disc will be sent to the cognizant discipline node of the Planetary Data System. For further information about the SDDPT and the process by which valuable data sets were identified, see the SDDPT Inventory/Evaluation Phase Final Report,(JPL Document D-10753) or contact: PDS Operator MS 525-389 Jet Propulsion Laboratory 4800 Oak Grove Drive Pasadena, CA 91109 818-306-6130 Electronic mail addresses: Internet: PDS_Operator@archive.jpl.nasa.gov 2. Data Set Information and SDDPT History The input data for these CD volumes were originally archived on 7-track and 9-track magnetic tapes and contained several types of special Viking Lander archival products generated by the Image Processing Laboratory (IPL). The 7-track data was packed from 4-6 bit bytes into 3-8 bit bytes in variable length records. Packing format used by /PACK to convert four 6-bit inputs to three 8-bit output bytes: First six bits input go to upper six bits of first output byte. Upper two bits of second input go to lower two bits of first byte. Lower four bits of second input go to upper four bits of second byte. Upper four bits of third input go to lower four bits of second byte. Lower two bits of third input go to upper two bits of third byte. Six bits of fourth input go to lower six bits of third byte. Research indicated that the 7-track tape data was originally written in "snaked" format with attached ebcdic vicar labels in front of the data. THE 6 BIT CHARACTERS ARE PACKED TO 8 BITS BY UNSNAKING THE BITS, EXAMPLE BELOW. 7-TRACK ---> DISC FORMAT ============= =============== P D D D D D D \ D D D D D D D D P X X X X D D \ X X X X X X X X P Z Z X X X X / Z Z Z Z Z Z Z Z P Z Z Z Z Z Z / WHERE P IS THE PARITY BIT AND D,X,Z ARE DATA BITS THAT COMPOSE 8 BIT BYTES. The 9-track data files are comprised of fixed length records - for more detail on the exact file structure, consult the COPY_LOG file accompanying each tape. The file VL_INFO.TXT is located on each CD in the series. It correlates the tape ID with available descriptions of the tapes' contents and the tracks. The tape descriptions and tracks were obtained from the exterior tape labels and available information contained in the MIPS On-line tape database. More detailed descriptions may be available from the analysts' catalogs. Some tapes were found to have incorrect track information during processing. Please consult the VL_INFO.TXT on the last CD volume of this dataset for accurate information on tracks, additionally the archive CD volume is also included for each tape. In order for the user to locate a particular file on the appropriate CD volume, he or she can check the jewel box backliners to determine the input tapes used to create the particular CD, or consult the VL_INFO.TXT file on the last CD volume in the series. 2.1. Displaying Images To display images originally archived on 7-track tapes use the VICAR program LABEL-CREATE, with inputs of NL = xxx (number of lines) and NS = xxx (number of samples). To display images originally archived on 9-track tapes use the VICAR program DCONVIM with input = record length. 2.2. Missing Tapes Some of the VL tapes were known to be missing at the time these archive CDs were created. They have been identified in the VL_INFO.TXT file. 3. Disc Format Each CD-ROM disc has been formatted so that a variety of computer systems (e.g. IBM PC, Macintosh, Sun, VAX) may access the data. Specifically, the discs are formatted according to the ISO 9660 level 1 Interchange Standard, and file attributes are specified by Extended Attribute Records (XARs). For computer software that fully supports XARs, access to the CD-ROM volume will be straightforward; the disc will appear to the user to be identical to a file system of directories, subdirectories and data files. Some computer systems that do not support XARs will ignore them; others will append the XAR to the beginning of the file, in which case the user must read past the first 512 bytes of the file. For further information, refer to the ISO 9660 Standard Document: RF# ISO 9660-1998, 15 April 1988. 4. CD-ROM Contents The files on this CD-ROM are organized in one top-level directory with several subdirectories. The following table shows the structure and contents of these directories. In the table, directory names are enclosed in square brackets ([]), upper-case letters indicate an actual directory or filename and lower-case letters indicate the general form of the directory file names. Top-level directory | |-AAREADME.TXT - This text file containing general information. | |-[INDEX] - Directory containing a text file to simplify searching for | | specific data. | | | |-VL_INFO.TXT- Text files describing basic information about the input | | data. In these files, tape IDs are correlated with their | descriptions. On the last CD of the series, the CD volume | names are also provided. | |-[VL] - The tapes are grouped into directories based on prefix | | sets. This is done in order to minimize delays reading | | the CD due to a large number of directories all appearing at | | the same level of the CD. In this example, JS is the | | prefix of all the tapes included in the directory. | | | |-[VL_xxxx] - The data files are arranged into directories based on the | | | INPUT TAPE IDS. The user can locate particular data based on | | | the tape's description by using the VL_INFO.TXT file on the | | | last CD of the series. | | | | | |-[DATA] - The subdirectory containing the data files. | | | | | | | |-VL_xxxx.yyy;1 - Each data file has a unique name. The file | | | name (VL_xxxx), represents the input tape id, and the file | | | extension (yyy) indicates the file number on the input tape. | | | For example, VL_0067.001 is the file originally located on | | | tape VL_0067, file 1. | | | | | |-[LOGS] - The subdirectory containing a log file describing tape | | | | copying history. | | | | | | | |- COPY_LOG.yyy;1 - The tape copying history has been preserved for | | | each data file so that the user can determine if some | | | data has been skipped or lost due to problems reading the | | | input tape. There will be a COPY_LOG in the directory | | | covering each file on the input tape. If there is only one | | | log file in the directory, it covers all files on the tape, | | | otherwise the file extension (yyy) indicates the file number | | | on the input tape. If problems were encountered during the | | | copying process, the entire history will be preserved in | | | this log file. 5. Text Files All document files and copying logs contain a carriage return character (ASCII 13) and a line feed character (ASCII 10) at the end of each record. This allows the files to be read by the MacOS, DOS, Unix, and VMS operating systems. 6. Recommended CD-ROM Drives and Driver Software (Note that the following list is not inclusive.) VAX/VMS Drive: Digital Equipment Corporation (DEC) RRD40, RRD42, or RRD50. Driver: DEC VFS CD-ROM driver V4.7 or V5.2 and up. Note: The driver software may be obtained from Jason Hyon at JPL (Files are: JPLPDS::DISK$USER1:[JHYON]VFS*.A). It is necessary to use this driver to access the XARs on the CD-ROM. VAX/Ultrix Drive: DEC RRD40, RRD42, or RRD50. Driver: Supplied with Ultrix 3.1. Note: Use the "cdio" software package (in "~ftp/src/cdio.shar" from the "space.mit.edu" server). IBM PC Drive: Toshiba, Hitachi, Sony, or compatible. Driver: Microsoft MSCDEX version 2.2. Note: The newest version of MSCDEX (released in February 1990) is generally available. Apple Macintosh Drive: Apple CD SC (Sony) or Toshiba. Driver: Apple CD-ROM driver. Note: The Toshiba drive requires a separate driver, which may be obtained from Toshiba. Sun Micro Drive: SUN Microsystems (Sony). Driver: SUN CD-ROM Driver. Note: There is a patch to support structured files. >the following error, when >trying to do a "ls" on its directory contents: > 'hsfs: filetype(0X8) not supported'. > The files that cause the error message are structured files. They are organized in records (fix or variable length). Applications need to retrieve additional file attributes (fix or variable length, and maximum record length) from these files in order to access the contents correctly. Unfortunately, UNIX can only handle stream files (unstructured byte stream). There is currently no application interface to access the extended file attributes and data records in a record format file. (Note: Philips is coming up with an X/Open specification (XCDR) to provide an API to access the ISO 9660 CD-ROM format Disc. It is too early to tell whether it will be approved or not). The design decision when implementing the CD-ROM file system was to ignore all record format files stored in a CD-ROM. This is obviously a mistake. The best way is at least to allow users to access these files. The following is a patch for SunOS 4.1 Sun4c (SPARCSTATION) to correct the above problem: acuraintegra# adb -w vmunix >>>> hs_parsedir+9c?X _hs_parsedir+0x9c: ba0da0ee >>>> hs_parsedir+9c?Wba0da0e6 _hs_parsedir+0x9c: 0xba0da0ee = 0xba0da0e6 >>>> hs_parsedir+14c?X _hs_parsedir+0x14c: ba0da0ee >>>> hs_parsedir+14c?Wba0da0e6 _hs_parsedir+0x14c: 0xba0da0ee = 0xba0da0e6 >>>> $q Note: you can also patch hsfs_node.o with the above patch. The patch for SunOS 4.0.3c (SPARCSTATION) is similar: hs_parsedir+9c?X _hs_parsedir+0x9c: ba0de0ee hs_parsedir+9c?Wba0de0e6 _hs_parsedir+0x9c: 0xba0de0ee = 0xba0de0e6 hs_parsedir+14c?X _hs_parsedir+0x14c: ba0de0ee hs_parsedir+14c?Wba0de0e6 _hs_parsedir+0x14c: 0xba0de0ee = 0xba0de0e6 $q The following is the patch for the sun3: adb -w hsfs_node.o hs_parsedir+56?X _hs_parsedir+0x56: ee6612 hs_parsedir+56?We66612 _hs_parsedir+0x56: 0xee6612 = 0xe66612 hs_parsedir+e4?X _hs_parsedir+0xe4: ee6612 hs_parsedir+e4?We66612 _hs_parsedir+0xe4: 0xee6612 = 0xe66612 $q 6. Who to Contact for Information For questions about how to read CD-ROMs: Jason J. Hyon MS 525-3610 Jet Propulsion Laboratory 4800 Oak Grove Drive Pasadena, CA 91109 818-306-6054 Electronic mail addresses: SPAN: JPLPDS::JHYON Internet: JHYON@jplpds.jpl.nasa.gov NASAmail: jhyon For questions concerning detailed information about the Viking Lander Processed Archival Images (TDRs) contact: Sue LaVoie MS 168-514 Jet Propulsion Laboratory 4800 Oak Grove Drive Pasadena, CA 91109 818-354-5677 Electronic mail addresses: Internet: Sue_LaVoie@iplmail.jpl.nasa.gov For questions concerning the distribution of the Viking Lander Processed Archival Images (TDRs) contact the Planetary Data System's Geosciences Node: Ed Guinness PDS Geosciences Node Washington University, Campus Box 1169 One Brookings Drive St. Louis, MO 63130-4899 314-935-5493 Electronic mail Addresses: E-mail- JPLPDS::EGUINNESS TCP- guinness@wunder.wustl.edu See also Ray Arvidson.