This topic is for the purpose of discussing #435: Mana.'s NES Super Mario Bros. "warps" in 05:01.27

This is an automatically posted message for discussing submission: #435: Mana.'s NES Super Mario Bros in 05:01.27 See details at: <a href="http://tasvideos.org/435S.html">http://tasvideos.org/435S.html</a> Download submission here: <a href="http://tasvideos.org//submissions/435.zip">http://tasvideos.org//submissions/435.zip</a> Submission status: <a href="http://tasvideos.org/242M.html">published</a> Last edited by: Ilari

---

The submission message follows: <blockquote class="quote">I received this submission in email from mr. Mana's friend, mr. Pom. <p/>It was made using VirtuaNES 0.92, but I converted it to Famtasia format using <a href="http://tasvideos.org/Blip.html">blip</a>'s conversion utility. <p/>The movie originally had a 72-frame wait at the startup screen and stopped the movie a little late, because the author believed that only the time between pressing Start and at the end matters. However, because we need to be able to calculate and compare times precisely, I updated the movie to eliminate these delays. This required me to replay some parts of the movie due to changed enemy paths. The parts were: <ul><li> 8-2: time=354 </li><li> 8-3: time=296 </li><li> 8-3: time=293 </li><li> 8-3: time=279-275 </li><li> 8-3: time=269-250 </li><li> 8-4: time=315 </li></ul>Regardless of these changes, I still consider this Mana.'s work. <p/>This is a full completion of SMB and aims for speed and entertainment. It does not have any other mathematical goal (such as best score). <p/>Mr. Mana. is Japanese and not really skilled in English language, but he is registered at the nesvideos forums. <p/>-- <a href="http://tasvideos.org/Bisqwit.html">Bisqwit</a> </blockquote>

I managed to remove those delays (I had to replay a few small parts of 8-2, 8-3 and 8-4 to adjust for the changed randomness). The length is now 05:01.27 (18076 frames). I'm waiting for Mana's permission to use that file.

Wow. I didn't think anyone would beat SMB any faster. How did he runs through walls like that?

Those collision checks... :P

Hmmm, I don't seem to have the Japanese version of the rom, nor can I find it... so I encourage the making of the avi files ASAP!

With the delays edited out, this is a must-publish. :)

Interesting.. managed to play the movie back in my emulator without any desyncs (which has far better audio than Famtasia). Perhaps Nintendulator could make another minor cameo when this movie gets published :).

Wow, those glitches were very clever and perfectly used. What a great video. It's so good to see that even the videos that seem 100% flawless can still be improved.

Obviously you can reach the vine in 4-2 without slowing down, as Michael Fried does, yet he (must?) backtrack to take the pipe up the vine instead of climbing. Does anyone know why what is done in the new video allows the pipe to be taken without backtracking? Why is backtracking needed in Michael Fried's case? Is it possible to combine the two strategies and not slow down and not backtrack?

Obviously you can reach the vine in 4-2 without slowing down, as Michael Fried does, yet he (must?) backtrack to take the pipe up the vine instead of climbing.

The idea of this glitch is to get Mario ahead of the normal screen scrolling position. Michael Fried did this by using the wrapping glitch (move from screen left side to right side by using the vine as a tool). Mana. did this when he entered the brick wall - during the first ~16 pixels of movement, the screen did not scroll. Apparently this was enough. The same thing happened in 1-2, but there the effect was against him - it would have been beneficial to have the screen scroll ahead of Mario, because that's what triggers the warp zone texts.

I voted no because there were a few places where he lost a few frames so it's not perfect, and he also missed some coins and made some jumps higher than they had to be. If I find the time I'll probably make a new version myself and make it completely perfect.

Wow I can't wait until the fixed version of this is made.

Why should your video stay and his video not get published just because he missed a few coins and didn't do things absolutely perfect. Yes, I know the goal is absolute perfection. I am well aware of that. My point is that he comes a whole lot closer to perfection than your video does, Michael. Someone happened to improve on your video which you thought to be perfect. I understand this SMB of yours is your pride and joy, but someone shattered your time. Congratulate him and then get to work on achieving perfection instead of talking about it.

Well, I think that it is worthy of being published, but I think that it shouldn't be because of two reasons. Like Michael Fried said, he missed some coins and coin blocks (don't know about the lost frames), and secondly because Phil and now Michael are both making improved runs. Too bad that none of the new tricks can be used in the warpless run.

Right before he posted, I said in IRC what mikebastard did in his post. I agree(and others did too). The only reason I see why it shouldn't immediately is because as bob said 2 others are working on it. But I don't know how long that will take anyway.

Too bad that none of the new tricks can be used in the warpless run.

At the end of 2-1there just might be that.

Fried said if he can find time then he will probably make a new perfect version. Unless either him or Phil can guarantee a new version within a few days then I think Bisqwit should go ahead and publish this one, if he gets the okay that is. It shouldn't be too much work for Bisqwit seeing how the file will be pretty small.

I have already encoded the optimized version but I'm waiting for Mana's comment at it.

Why should your video stay and his video not get published just because he missed a few coins and didn't do things absolutely perfect. Yes, I know the goal is absolute perfection. I am well aware of that. My point is that he comes a whole lot closer to perfection than your video does, Michael. Because some parts of his video were significantly worse than my video, and I noticed the mistakes the first time I watched it while watching it at full speed. It could probably be over a second faster by just fixing the inaccuracies. If he's gonna add in a new glitch to make it faster, he should at least keep the rest of it as good. Edit: Fried said if he can find time then he will probably make a new perfect version. Unless either him or Phil can guarantee a new version within a few days then I think Bisqwit should go ahead and publish this one, if he gets the okay that is. It shouldn't be too much work for Bisqwit seeing how the file will be pretty small. I'll try to get it done within a few days, but I can't make any promises because it all depends on how much time I decide to spend playing Mario Kart 64 and also how much homework I get. Btw, before I make too much of my run, does anyone know if Phil found any new strategies or is he just gonna make Mana's run more accurate?

I voted no because there were a few places where he lost a few frames so it's not perfect, and he also missed some coins and made some jumps higher than they had to be. If I find the time I'll probably make a new version myself and make it completely perfect.

If "it can be improved" would be reason enough not to publish, then there probably wouldn't be any published movies at all really. ;) And missing coins doesn't matter - he aims for the fastest time. Everything else is just an added bonus.

If coins can be collected without sacrificing time, do it(though it's no real reason to reject a run. I know that's not the only reason in this case). You have nothing to lose from doing so. Phil said, and I quote, that he's found some "top secret" trick he plans on using, and plans on breaking 5 minutes.

Well, maybe a few coins aren't enough, but it's policy here to reject a movie if it is faster by a little bit but played more sloppy.

If "it can be improved" would be reason enough not to publish, then there probably wouldn't be any published movies at all really. ;) And missing coins doesn't matter - he aims for the fastest time. Everything else is just an added bonus. It wasn't just that it can be improved, it was that he did some parts slower than I did. Also, there were coins he could have gotten that wouldn't have cost him any time. Btw, I finished the first two levels of my run. After 1-1, I'm ahead by 84 frames, mostly because he took awhile to press start, but I think he also lost a few frames in the coin room. After 1-2, I'm ahead by 85 frames, but I'm not sure if I did the last part perfectly so I'll see if I can come up with a way to improve it. Edit: I redid the warp zone part and now I'm 90 frames ahead after 1-2.