Spirals corruption
The Spirals corruption is a glitch in some versions of Chip's Challenge which concerns an alternate version of Spirals with one small difference. At grid coordinate [29, 13] is a thin wall instead of a floor space, which causes a much harder level worthy of a 5 rating.
The cause of this alternate level is theorized by Ruben Spaans to be a result of the conversion between MS and Lynx, which caused the Transparency Glitch, boosting and Controller and Boss Glitch, among with most of the other current glitches known. The programmers appear to have done very little playtesting in the MS CC emulation, and found that Spirals was exceedingly difficult because of the different rules of walker movement in MS. Although they edited the level, some copies of Spirals, generally found in Windows Entertainment Pack 4 versions of the MS Chip's Challenge, never received the edit. All copies of the Commodore 64 emulation also contain the corrupt Spirals.
This corruption was first brought to the full attention of the public, although isolated reports had been made, on March 23, 1998, in a discovery by Erik Arfeuille:
There definitely was a bug in my chips data file. I used the chip's editor and the instructions given to me by Richard Field [see the link to his site in Other Fan(atic)s] and was able to remedy the problem. If you have any knowledge of the Chip's editor: run it, load level 88 and check out square 29,13. On my Level 88, this square was blocked to the east by a thin wall! Level unsolvable. Via Richard Field's Chip's site, I learned that I wasn't the only one with this problem. Nobody knows why only some Chips.data files are corrupted in this way.
Ruben Spaans attempted the corrupted level, and found it could indeed be solved with a Texas-size portion of luck, reporting a time of 300. The full text of his theory is quoted as follows:
...I believe [my "corrupted" Chips.dat] is not corrupted. The only difference between [it]and the ["normal"] one ... the blocked spot above the exit on level 88. I did a file compare, and this is the only difference... What's makes this a bit interesting, is that level 88 on both the Lynx and Commodore 64 is exactly like my "corrupted" version... Somehow the walkers aren't as aggressive on the C64 as on the PC, and therefore level 88 isn't at all impossible on the C64 (I haven't played the Lynx version, so I can't speak for that one). My theory is that Microsoft converted all the levels from the Lynx, with too [little] playtesting. Later they discovered that level 88 was way too difficult, so they patched it. Surely you know about all those non-working trap buttons and clone machines, so apparently they slipped up on many levels. But that's just my theory about level 88.... For the technical minded, the level 88 difference in [is?] at byte offset $EC5C in the Chips.dat file. The blocked square version ("corrupted") contains the byte $09, while the open square has the byte $00. So everyone with a hex editor can fix their file.
Since this time, the corrupted version's bold has reached 304, achieved with turning 3D U 2D 4R 3L at the intersections, and with at most three moves lost. Though still significant, the luck needed is not tremendous, particularly if the player is a bit mindful of how many walkers are created in the first moments of the level. If they see less than ten walkers before picking up the second computer chip, then their chances of at least completing the level are decent. As for fixing a corrupt level, instead of using a hex editor, edit the level in a level editor.
Sources
Not All The Bugs Are Yellow: Alice Voith's analysis of the then-known glitches
Ruben's score report and 300 solution, posted on the Charter Chipsters forum