This topic is for the purpose of discussing #3668: WinEpic's DS New Super Mario Bros. "No Cannons" in 57:00.46

a warpless run? why not

i'd like an encode if possible thanks.

Is it just me, or is this run slower (in the levels both runs complete) then the published run? Edit: Changed the wording. I am a native English speaker and even I make mistakes. It happens.

Funny. I was literally thinking yesterday (What if someone did a warpless run on NSMB on DS?) Wow, that was the most unexpected. Definite yes vote.

I feel a little concerned when the author writes, "I could probably improve my run by a few minutes..." A few minutes is a lot of time. Wouldn't it be better to keep it on the bench until you feel you've optimized it as much as possible? There's no rush to get something published.

Voting no for obvious reasons. The idea is great and all, but if you're behind the published run by a whopping 4 in-game seconds at the end of the second level... It's not exactly something I want to see published. :/

Reall bad optimized. Voted no.

Indeed, this would work much better as a WIP than as a submission.

I was giddy and excited to see a warpless run of this game. Then I actually watched it. Today was a sad day for this hellcat. A very sad day indeed.

Havent' watched it yet, I'll wait for the encode probably. BUt 1616 rerecords, that's like a 30 second TAS for me. And this is like what, 57 minutes? I'll hold my vote til' I see the encode.

I hope you will work on improving this. I want to see a good run of this.

Unoptimized, no vote. Also, 1616 rerecords is too low. You need to take your time and optimize a run like this before submitting it. I would be interested in an optimized run of this category, or even better, a warpless run.

I cannot comment on the optimization, but the "No Cannons"-choice rubs me the wrong way. Either do it as fast as possible, or do the full 100% run with all levels. Accepting one kind of level-skipping feature but not another just seems arbitrary. Waiting for an encode before voting.

Either do it as fast as possible, or do the full 100% run with all levels. Accepting one kind of level-skipping feature but not another just seems arbitrary.

Actually I've seen some examples like that which are pretty decent around the site. I don't really see how being just a bit arbitrary is wrong as long as it still has a logical reason to be accepted. An example of this can be found here http://tasvideos.org/2062M.html

By default, the categories are any%, 100% and (if breaking glitches exist) some kind of unglitched any%. SM64 "70 stars, no BLJ" is accepted for three reasons: * 70 stars is what the game demands, thus it has always been an accepted category in this game, making this the unglitched any% category. * in SM64 you can choose which 70 of the 120 stars to collect, with very few restrictions. Picking just the fastest stars is a good thing for an entertaining run. * BLJ wasn't banned to show more of the game, but because it looks ugly when abused too often. Also, it's the unglitched any%. The special category is justified. Most importantly, it's acceptable because it offers substantial differences to both the 100% and any% categories. In this NSMB DS, you don't get to pick the fastest levels, but essentially a random subset, and you end up with a total number of levels that bears no importance at all. I'd compare that situation to super mario world (and its hacks), where all existing runs are either low% or fully 100%. Unless a "No cannons"-run is significantly different from a 100% run (and I don't think it is), IMHO only one category should be accepted, and that's usually the 100% one because it's less arbitrary. Now you'll bring up the warpless SMB3 run. Again, I think the special category is ok there because (unless a true 100%), it can show off the item system and the extensive route planning such an item system entails. Again, I haven't watched this submission, so I may be misinformed about NSMB DS. My point is that a run on an unusual category should require some justification; the category choices should be made for the benefit of the run, and not just because the author felt like it.

-Typed on a 3DS btw- NSMB 'no cannons' is basically a SMB 3 warpless run. A warpless run of this game can only go through 6 out of 8 worlds. W3 vs W4 and W6 vs W7. A 100% run of this requires collecting and spending 240 star coins, clear all 80 stages, find all cannons and secret exits (W5-Castle secret exit is NOT required as both that exit and W4-cannon unlocks World 7. W4-cannon is higher priority than W5-Castle alt exit in terms of unlocking the 3 star next to the player's file (100%))

This is a valid category: unlike 100%, it doesn't require hunting for star coins. So it's seeing a reasonable number of the levels completed as fast as possible, something that doesn't exist in other categories. (It's thus a little more different from 100% than, say, SMB3 no-whistles, which is a very similar category.) I have a feeling that this run is unoptimised, though. My instant reaction simply upon seeing the time was "that feels too slow for TAS warpless", and other people's comments are making me feel that's right. (Waiting on an encode before I come to a real judgement on that.)

Encode downloadable from Sendspace.com (~235MB). Click on link in the blue box below the filesize. Don't click on any other buttons; those are ads. http://www.sendspace.com/file/oke2ns The actual time of the TAS should be around 51:42.

Didn't we already have a skip-entire-the-game run of this game?

While I do want to see a warpless run of this game, your run is not up to snuff. I would hope that of the levels you play that are the same levels the published run does, you should get the same or better times.

Desyncs for me after the first stage.

henke37, you need to enable "Advanced Bus-Level Timing" If it is enabled, are you using 0.9.8? What is the SHA1 of the ROM file? My hash is: A22713711B5CD58DFBAFC9688DADEA66C59888CE