Although I generally prefer not to express myself more than necessary on the subject -having neither taste nor the will to do so- i would like to develop some additional points and recollections.
I never played nor got interested specifically on FF6 until the middle of February. At this time, during a run of MV, i asked to myself "when does a character enter the memory of the game?". The only reasonable time being the time when the character is being named, I suggested to MV to try to semi-skip Celes. We tested it right away on a secondary stream -Synahel and other friends already being there at this point- and thanks to his knowledge we came to the now classical Locke's scenario strategy.
In spite of the fact that this improvement was not essential, I still thought of it as a valuable and esthetic artifice of the run. This drove me to the will of looking more closely to the route. I then encountered the 'Let's play archive' dedicated to FF6 glitches, and my inspiration was stirred by the description on how it remains possible to glitch the game 'a la japonaise' -through the 255th slot- using the sketch glitch. I began some tests and ended up with an abstract route, which essentially delayed one of the most intricate problem : the magic route.
The toying around thus swiftly turned into an empiric study of the game, and I discovered step by step through experience how the game deals with magic, not knowing that TASers had already unleashed those mechanics previously with their tools. I came up with the banal magic route going for the pseudo-mandatory Vanish magic, whose drawback -among many others- was to force an Intangir fight. I was not satisfied with it and i would have not accepted to release such rough route. At this time MV and Brandy were sometimes coming to my home and supported me, as i was exhausted struggling with all these datas.
For some reason, I once saw a reverse Phantom strat and discovered right after that it was actually possible to Sketch Glitch using Muddle, in stead of Vanish. This detail is one of the most important features of the Sketch Route, since it allows more flexibility in the magic learning. I then went again for empiric research on the shortest path to slot 28, and came up with a really fast one, consisting mainly on naturally learned Terra magics, as well as Muddle-Imp, which are 'mandatory' to trigger the glitch. The game was then routed and learned in some weeks, with the other new improvements and circonvolutions I wrote in my notes. I then came to Kilaye and offered to work with him on a TAS. He was very busy at this time and couldn't get into it.
Some weeks after -on the 28th of June, if i recall correctly- Kilaye came back to me and was now ready. The Skype conversation 'FF6 - constant brainstorm' was created and the 'French TASers', as people seem to call us, began to work. As i said here and there, more than 600 pages of conversation were needed to realize this TAS, and there has not been any line on which we didn't have the same point of view. At some point I was reluctant to starting TASing the begining of the game, fearing bad random encounter manipulation, but Kilaye decided that we should go in, even though the strategy was not settled. I would like to point out that Kilaye has been in charge of almost all the technical realization of the TAS. The admirable self-denial he demonstrated when solving all these micro-issues is impressing.
Simon and myself offered a first Wrong Warp route, which had lots of subtle circonvolutions but was satisfying in terms of swiftness. Kilaye thought that this could be improved, and worked on another way to Warp, using subtle tricks such as Tier changes. He came up with an hypothetic new way to Warp and for some reason, as soon as he explained this to me, I asked him "... nevertheless, why don't we do 'this'?". The discussion dropped instantly and 'this' became the route of the TAS, which in fact remains in the same spirit as the first route, being slightly faster and much safer.
One of the greatest personal satisfaction of all of this is that the final magic route of the TAS is actually the magic route i came up to empirically from the very begining. Simon has shown rigorously that this magic route is the fastest possible to Sketch Glitch.
There are three points on which the TAS did not eventually satisfy me completely. First, I still have the faith in a warp to credits (Kilaye will probably say that it's really hard to remove things, once when they get into my mind :)). Hopefully there will be some oppportunities to do so, and we stay aware of this. Esthetically, i would also have loved to farm for Magic Points using Reverse Magicite to cast Phoenix and instant kill Undead ennemies three by three, but the banal strategy using Rods appeared to be a bit faster. Finally, I would have loved to finish the game escaping from Tier 1 using Engulf, stop the timer at the fadeout just as we do in the speedrun of the game and loop back to Zone Eater's belly, getting again to Kefka's tower and loop back again and again during tier 1. Unfortunately, we have proved that it is not possible to do so.
As for the softlock during the ending, don't forget that this does not rely directly on memory corruption, nor on the way the final fight ends (indeed, the ends goes normally -even if you corrupt memory- as long as there is no Wrong Warp). Since the usual rule is to stop the timer at the last input, i don't really see in which way this is an issue. I would advise people who still feel roughly offended about it to look through a window, wait for a plane to appear in the sky and turn their head around sometimes to get the real ending. If people still feel deeply betrayed, the submission can be withdrawn.
This TAS thus owes a lot to Kilaye's tremendous work, and my feelings and mind are bounded to all those who accompanied us during this journey as much as they could, notably Simon, Brandy, MV and Synahel.
L.L.
PS: hopefully, the 'French TASers' will run new projects in the future.