In the 80s and 90s of the last century, backup systems on magnetic tape were very popular because of their relatively large capacity and affordable prices. At that time, many were convinced that the backup tape, which is stored somewhere in the far corner of the cabinet, will be in fair condition even after years, and it can be read.
Now technologies have gone far ahead, many have already forgotten about the magnetic tape, but the information stored on them may be of some interest. Therefore, there are enthusiasts who are looking for ways to count this information.
The Strand Games team, which this year was founded by Magnetic Scrolls co-founder Hugh Steers and interactive fiction fan Stefan Meier, this year set about restoring Magnetic Scrolls game code to revitalize text quests and create new open source Interactive Fiction game design tools. . The company has already released a restored version of 1985’s first Magnetic Scrolls game, The Pawn
The team then decided to restore the TK50 cartridges with the only known backup of the Magnetic Scrolls games developed on the DEC MicroVAX minicomputer. All backups were made by a central minicomputer using backup technology.
This work is associated with the search and recovery of significant achievements of the gaming industry, which often require a lot of time to preserve their historical value. In one of the cases, it was necessary to arrive in an original way - bake precious cartridges with the world's only copies of the source code of the early text quest of Magnetic Scrolls for about eight hours in a household oven at 45 °.
For a long time these cartridges were gathering dust on the shelves, and in the end they were put in boxes and forgot about them for seventeen years. Then Stirs got the idea to read and republish games on them. However, in 2000, I did not find suitable means for reading the tape format that was outdated at that time, and put the idea aside. The boxes continued to gather dust in the office until one of the employees brought them to his home, where they continued to lie in the attic.
After another seventeen years, by a happy coincidence, the cartridges were again at Stirs, and he decided to read them, no matter what. But how to do that? With this question, he turned to people on the forums, until he was thus acquainted with Rob Jarrat, who had a working DEC
computer and, which is also important, a compatible streamer in working condition. Rob offered to help recover the data. The problems seemed to be over: it’s enough to read the data.
Tapes are not read. It turned out that the old tape had lost its stickiness - she started the so-called sticky shed
syndrome. This meant that the tape in such a state blocked the streamer, could “hammer” the head and spoil the guide rollers. In addition, it was a bit sticky due to the fact that it absorbed moisture. In some cases, a bonding adhesive that holds iron (III) oxide in a plastic substrate may break, leaving rusty particles on the guide heads and, as a rule, on the surface of the tape drive.
Rob heard that some enthusiasts could solve it by baking. The idea was to heat the ribbon for several hours at a relatively low temperature for the oven to dry it and then read it without any problems. However, no one could say for sure at what temperature and how long the bake should have been baked. In addition, it was not known what would happen to the tape if accidentally overheated or overexposed.
They tried to find the right recipe. The first experiment involved an old ribbon, which was baked at 50 ° for 5 hours. Then Rob and Hugh came to the conclusion that the temperature was too high, and the tape could deteriorate. But since the state of the tape before the experiment was unclear, there was no certainty that it was due to the effects of high temperature that it deteriorated.
Then Rob tried to bake another test tape at 40 ° again for 5 hours. It almost helped. Some of the unreadable tape before baking was read, but then it got stuck in the drive. The same tape was left in the oven first for 8 hours at 40 °, and the team was able to restore the list of catalogs before the tape was stuck again. Finally, the tape spent another 8 hours in the oven at a temperature of 45 °, and it worked. The tape stopped sticking, and Rob and Hugh restored some actual files.
In the same way, baking solved the problem of stickiness of cartridges with Magnetic Scrolls games, but did not work to combat oxide deposits, due to which the tape became unreadable. Rob could clean the drive head every time it was read, but he wouldn’t allow the entire tape to run out before it was clogged again. This problem was aggravated by the fact that the TK50Z streamer scrolled the tape before reading in search of the necessary section. He needed to know what was on the tape before retrieving the data.
On the TK50Z drive, it is impossible to clean the heads when the tape is installed - it needs to be lifted to get to them. But when the drive is running, the tape cannot be moved. This is due to the fact that during operation, the heads move up and down, and the engine, which causes them to move, is turned on only when there is power.
Therefore, the TK50Z could only be cleaned after removing the tape. And each time another tape is loaded, the tape drive scrolls up and down as DLT tapes store data in parallel tracks along its entire length.
Then the team decided to abandon the TK50Z and switched to the TZ30, which can read (but not write) the TK50 cartridges. The drive itself did not particularly solve the problem, but the team found that it was possible to remove a pair of retaining rings from the TZ30 mechanism and raise the head enough to clean it with cotton swabs moistened with isopropyl alcohol. In this way, the actuator head could be cleaned before the actual data reading.
As a result, it was possible to read 70MB from one tape and 54MB on the other, while the maximum capacity of the cartridges was about 90MB. Both tapes had copies of important source data, and a comparison of the files recovered from each tape showed that they were identical and readable.
In their blog, the
team reported that the source code of several games was fully restored. Standard Games is currently working on a release of the remastered version of the 1987 Magnetic Scrolls game The Guild of Thieves, which will include the original source code and special thanks to Rob Jarrat.