Thanks for the information, it is much appreciated. Hopefully now some Game Gear TASes can really be produced, this is a very good news :)
btw, why weren't those changes documented in the Dega forum? How could anyone know that it had been fixed?
As much as I love the Game Gear, I should remember you that Dega has been declared inaccurate in that regard, by using 9-bit color instead of 12-bit color. It might be more or less noticeable according to the specific game, though. While most of the Master System ports don't make use of the extended palette, most of the Game Gear only games do.
Also the mmv format "needs" (lol) a new flag which will be set to 1 if the emulator decides to start in Game Gear mode by looking at the rom extension.
If I remember correctly, 15 stars are needed to trigger the rabbit which was used to get through the DDD door, and one star is collected in DDD to move the liquid wall and get access to the pit that leads you to Bowser. This, until it has been found that you can skip both of those tasks by BLJing through both the door and the liquid wall. However, and still, if I remember correctly, BLJs are needed to complete a 16 stars run too. The only difference between a 16 stars run and a 0 stars run is that there is one more BLJ.
I think I have to trust Baxter (the Tetris master ;)) about the 7 line clears here and thus I'll probably vote meh on this.
I, instead, demand evidence of that. Because unless proven otherwise this is the fastest completion currently available and therefore should be accepted as such. I've seen runs published where improvements were known, now you're suggesting this should be rejected because improvements may exist. Why do I see double standards implied in your words? I don't know if luck could be manipulated better, but I know that I couldn't delete any more rows at the same time, because each time I tried I died, so I backtracked and deleted fewer lines until I could survive. I spent several rerecords to do this so I seriously think this is the fastest I can get.
MUGG wrote:
Good luck next time
Next time? Are we at this point already? This is not rejected yet. We all know it will be (despite the poll is saying that some people liked this run), but for now it's not, so please don't talk like that.
Last time I checked you were in Parsely Woods, you had most of the job done, there were so few levels left. I hope someone can finish this... I'd do that but I don't want this excellent run to be rejected :\
My first reaction when I saw its duration in the title was "oh no this is going to get boring after 5 minutes". Well, I was wrong. This was entertaining for the whole hour, and that's rare for submissions this long, they usually get boring and/or repetitive after a while, but not this one, it could retain my attention from the first to the last minute. Yes vote.
I can't reply to 3 and 4, but:
1) The previous version of this TAS (the one I told you about on irc) was improvable, yes. However I redid it because the first incarnation of this run was made with FBA 0.0.1 and it didn't sync with FBA 0.0.3; I tried to hex edit the old run into the new one, but I couldn't, since the fbm format is compressed, so I had to redo the whole TAS from scratch. It got 1 second and 20 centiseconds faster in the process, with better luck manipulation at the beginning (which was my main complaint in my previous run) and better played levels (which wasn't expected, to be honest). So this is not unoptimised for this car.
2) The unassisted 3'10" is achieved with the truck. The fastest unassisted run on that website with this car is 3'26", 10 seconds slower than mine.
3) No, I didn't know. I just took an all-around car ignoring that the truck was better at this purpose, and it can exploit the collision bug at the beginning better (I saw the unassisted movie on that site, the truck can get above 400 km/h while I could get only 300 km/h)
4) Yes, there is no reason. I admit that.
This is not an half-assed attempt. It beats the game as fast as possible. With a different kind of car. There is no point in the guidelines that says that I can't play with a suboptimal character (or vehicle in this case).
Yes.
I'm not going to redo the run with the truck. But I agree that if someone does, it should obsolete this one (if this game isn't given 8 categories).
In games where you can pick a character of your choice it's not uncommon to see multiple TASes published alongside, like it happens for Super Mario Bros 2 or Castlevania. Ideally there might be up to 8 movies for this game, as long as they're done by people without "nineko" in their screen names.
Link to that page please, I want to see what car he used.
edit: found that page: http://replay.marpirc.net/r/overtop
Unless I'm mistaken, the fastest record for the "Dominator" (which is the car I used) is 3'26"24. The other records refer to other cars. I never said I picked the fastest car. The car I picked is an arbitrary choice.
Add "plays with a suboptimal car" in the categories, I guess.
edit2: confirmed. The 3'10"21 has been played with the truck.
I wonder why there must always be someone who tries to throw shit on the TASes I make. When I said in irc that there is a bias against me I wasn't joking. When other people make TASes there isn't a whole storm of detectives trying to find errors at any cost (even inventing them like in this case). I'm tired of this.
Ok, thank you :)
I started the upload and I got asleep with the computer on, lol.
Dromiceius wrote:
Also, shouldn't this have been played on level (i.e., difficulty) 8? I don't remember exactly how it worked when I released 0.0.2, but I thought FBA would allow you to access the dipswitch screen?
Yes, it should have been played in level 8 if this would have made it more interesting, but the only difference between level 4 and level 8 in this game is that level 8 gives you less time to beat the levels, and I'm way faster than that limit anyway, so level 1 and level 8 are the same deal for me, since the other cars' behaviour isn't modified. Now that I think of it, a level 8 run would probably be a few frames faster than this in real-time because of the fewer seconds added between each level. But I aim for in-game time.
Here is an encode of this run. I'd add it to the submission itself but I fail at editing pages. I will edit this post eventually with a Youtube link, though.
Question. How do I hex edit a movie file? Apparently it's compressed :(
edit: nevermind, I redid the whole movie from scratch, it got one second faster.
So, I have 350 posts now, do I have to set up variable avatars or can I just get along with my usual one? I wouldn't know what kind of variations I can do to this...
One thing, are you sure you want to collect 100 rings in Gigalopolis 1? That zone is extremely straight, it can be speeded through very quickly using the shortcut. Hitting two ring monitors would give you 20 rings that, while not optimal, would result in a decent speed bonus (909).
Gigalopolis 2 on the other hand seems perfect for that.