Posts for hanzou


1 2
5 6 7
12 13
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
Side note (and double-post!)
KDR_11k in the submission topic wrote:
Hm, the ending's kind of a bummer, I think such a run should pick up the lv3 sword but not use it until Ganon appears.
You're not alone in this opinion, but I think the majority would still prefer the run to be swordless rather than 99% sword-abstinent. In retrospect, my idea for a cheater ending (one of several options proposed in the off-topic discussion of the Outlands topic) seems a bit absurd for a publishable movie. By comparison, even the use of a Sonic Advance 1 cam-hack is difficult to agree upon even though that serves a much more viable purpose. However, I still wanted to somehow see the game "beaten" swordless, so I went ahead and made this alternate ending for my own entertainment, using a savestate from Josh's first run: Zelda 1 swordless director's cut (2 MB) I used a RAM patch code instead of a Game Genie-like patch. This was much easier to find although not as elegant because it requires the code (00AD:002) be turned on only at a certain point rather than throughout the whole run.
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
Oh right. Silly question of mine about the Zols. I'm not sure what you're saying about 4 bombs vs 7. Around frame 8850 you left a bomb set sitting on the left side of the room. Walking over to pick that up would take less time than the extra walking and waiting required for the Octorok.
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
WIP looks good most of the way up to level 3, because you managed to fit some good damage boosts in there. A couple comments: I don't fully understand how drops work in this game, but I don't think you should have stopped to kill the Octorok if you can get bombs from red Darknuts. Also, why did you kill some Zols when you could have walked right past them? Keep up the good work!
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
Hey Josh. First of all I'd like to say I'm glad you're not giving up on this run.
Josh wrote:
I'm pondering a choice of two different starting paths. The most obvious one is this: *Power Bracelet *Up+A *Warp to upper right quadrant *Down, right, up, right, right, up, get 100 Rupees
Getting the power bracelet and 100 rupies way up northeast stuck out to me as an unnecessarily long way to start out the run. I've reached Gannon swordless on a console before, but I didn't need to get the Power Bracelet early and I don't see why you should. Have you considered the following basic route? *Go northeast and get 30 rupies from an Armos secret. *Up+A *Buy bombs for 20 *Beat level 1 (edit: 3?) edit: Realizing how long this topic is, I figured it might be better to go back and read some posts. Now I see Bob Whoops already assumed this route. So... isn't it faster?
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
Okay, after reading through a few times I think I finally understand all the issues here, and I can see why much confusion arose. There are just so many intricate issues when making a run for such a complex glitchy game. SmashManiac, I think you would have gotten a much better response had you stated your restrictions like so: Instead of saying no use of turbo, you should have said no use of the floating glitch. I think that's the real issue, because using that shortcut obviates many other entertaining glitches like the bouncing part. Restricting to no left+right is fine, and is even a bonus for the reasons KDR_11k stated. In addition, I think the wobbling trick requires left+right, so such a run would please viewers who find the wobble-walk annoying. There are specific guidelines for runs on this site, and even Bisqwit's walking run follows these for the most part. An important one is to stay within well-defined goals. Forbidding 30 Hz button presses is not a well-defined goal because it is an arbitrary restriction on player reflexes. However, the goal of finishing the game as fast as possible without left+right and without the floating glitch is clearly defined. We already have several runs on the site that are done with restrictions of not abusing specific bugs if those particular glitches subtract entertainment value. Even flagitious' movie is an example of such. In this sense, a frame-perfected version of your movie could even fit as a regular run rather than a concept demo. That said, as to whether such a movie would be accepted I am uncertain. There is a conscious effort to avoid having too many runs of the same game, and that might make it difficult to accept a third movie. From the looks of it though, I think I would vote yes if you made a frame-perfect run using this glitchy route.
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
Warp wrote:
unnecessary jumping
Jumping is faster than walking.
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
Sfida wrote:
why bother *sarcasm*
Ipher's site wrote:
Feel free to make your own Windows port from the source, but if you do PLEASE send the changes back to the SNES9x team.
Putting the word please in all caps suggests the authors want a Windows version just as much as we do. It's a huge amount of work though, and noone has all the time in the world. The same could be said of all of us here. We have available the old Windows + Linux source and the new Linux source. What's to stop anyone from diffing the sources and getting started right away on developing a Windows port?
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
You have to load from one of the savestates Shinryuu posted.
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
I really think manually making .srt subtitle files for these movies is the best solution to please everyone. Reasons: - I can't keep track of everything you guys have said about genuine shelf games, hack cheat worries, or updated translations being released, but this solves all of those problems. - Subtitles can be made for games for which we already have TAS movies in Japanese, rather than asking the author to redo the run. - Making subtitles takes time, but less time than it takes to make a good TAS. - Anyone familiar with the game--such as the author, fans of the TAS, or a judge/encoder person willing donate his/her time--can write these subtitles. - Writing subtitles for a TAS of a Japanese video game is generally easier than writing something like an anime fansub. Instead of having to think up your own natural-sounding words, you can just use already written lines from your favorite translation hack, or an official translation as nitsuja pointed out does exist for some games. - Furthermore, by using the framecounts in the emulator it's easy to estimate the proper timestamp for each line without going back to repeatedly test how it looks. - After publication, the subtitles can be iteratively revised (by anyone) without affecting the actual TAS. - When a movie is obsoleted, it takes little effort to repair the subtitles because usually time is saved in distinct spots so we can just timeshift those parts. - Instead of pausing or using slowmode to read the dialogue, a viewer can have the option of manually reading the .srt text.
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
Highness wrote:
Can you save any time by doing that? ...is it faster than the currently published material?
My guess is that this is a proof-of-concept that Quickman can be rapidly finished off with the air shooter, and the fight need not be done entirely with it. It was important, and difficult, to test whether or not Quickman could actually be manipulated for this strategy to work. Reason: For a while it has been known that the new run will have 2 or 3 extra crash bombs upon reaching the Quickman rematch. The potential here is to retain the same boss order, use bombs instead of P on Quickman, finish him with the air shooter (weapon ready for Crashman in the following fight), and thus save time while retaining the same number of weapon switches as in the published movie. Hope that made sense.
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
You could run the glitchy route in Megaman 2. This would involve playing all the way through most of the 'overflow' levels. The glitch combinations which I refer to as Heatman's volcano, Airman's snow mountain, Woodman's swamp, etc are all reasonably entertaining. Even when done alone as fast as possible this route would take about 35 minutes, so timewise it fits with the other 4 games. Megaman 1 might also be an option if it has overflow levels that are sufficiently colorful. Of course, six games would be quite a handful so then you'd probably want to drop one of the others.
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
I watched the Wily fortress part of finalfighter's run just now, and I spot only two places with even a remote chance of providing worthwhile refills. The first is in Wily2, where finalfighter uses item3 briefly. Using item1 might both be faster and give an opportunity for a refill just after killing an enemy with woodshield. The second spot is in Wily3, where finalfighter takes damage. Part of the time lost from an extra weapon switch would be made up for by not getting hit. All other refill opportunities seem to require at least two extra weapon switches, which is unlikely to pay off.
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
hopper: +5 insightful Tombad and KDR_11k: -1 redundant Oh, and good job there Fabian! Though I do wonder if the 96-exit run will ever be finished at this rate. So can you send me a preview of your Vanilla Dome 3? That should keep me happy for the time being. Thanks.
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
This can be done on all 8 robot masters in the original Rockman 2. It's just easier to pull off in Megaman Ultra because there's often more lag during boss fights.
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
For those of you who missed all the 'pit trick' discussion and experimentation 12 pages back: crashman_pit.fcm Scrollerwily2-Shinryuu.fcm Quickman and Wily2 were the only two levels where we found a way to survive (and save time with) the pit trick. Unfortunately, each of these instances fails to completely make up for frames lost in the boss fight. edit:
Bisqwit wrote:
Removing that strategy (pit trick) from the movie, we're 33 frames slower.
To what extent are you removing this strategy from the movie? Are you simply not turning left at the end, so as to avoid falling in the pit hence you don't shorten the teleport delay? By that time, you've already lost time having to rise a step higher on item1 and jumping across the shutter. Some of this is still offset by zipping to Bubbleman instead of letting a metal blade travel across the screen, but is this worth it? Have you tried kolechovski's suggestion? Fight normally then hit Rockman's head on the ceiling just before Bubbleman dies. I have wondered whether you can pull this off without losing any time in the fight.
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
I've already expressed my opinion that this game doesn't make a good TAS, worse so than Street Fighter or Mortal Kombat. Please watch Fultz Vol 1 if you have not already. One problem with a TAS as compared to a traditional combo movie is that you will have to watch all boring steps of the 80-hit combos (yes, 80 hits can be done on the SNES version). edit: Also, in order to perform the longest (excluding the 80-hit glitch) combos, you need to let the clock run down to zero, which is unacceptable in a TAS.
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
Here are a few links to explain some of the things mentioned so far. why you should jump down the chute at frame 2140 & 12300 : falling faster why Rockman 3 (J) instead of Mega Man 3 (U): avoids ~100 frames copyright screen (I can't find the thread where this was explained) why your Magnet Man fight was suboptimal: know for how long the boss' invulnerability lasts All in all, you've shown a strong effort in this but probably missed some of the necessary techniques because you compared your frame counts only at the completion of each stage. It takes a lot of work, but you instead need to compare times for every piece of the run where you potentially could lose frames. If you hold yourself to that standard you'll gradually come to understand how to make all parts of your run as fast or faster than Vatchern's movie. As for the 2nd controller cheat, I doubt there is much demand for a run utilizing the high-jump, since it'd only be a couple minutes faster while less entertaining and less impressive. I'd recommend trying to improve the regular Rockman 3 movie, except I'm not sure how far along Vatchern and Angerfist are with the next version. Would you be interested in Mega Man 4 or 6? Those runs are also visibly improvable and I think nobody is working on them at the moment.
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
Heh. I also once overanalyzed that idea, and probably others here have also spent too much time doing so. Have you thought about the stage select screen too? Fortunately (for sake of our heads), there aren't any pre-Wily spots in the run that involve switching directly between the air shooter and bubble lead.
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
Looks good. As for whether to go back and redo these four stages, the decision is up to you guys. I would suggest thinking about this in terms of which pieces you might successfully copypaste and which pieces will definitely need to be redone. Parts subject to randomness: - Woodman lag-prone area with apes - Crashman 1st substituted scrolling (relatively easy compared to 2nd?) - Crashman 2nd substituted scrolling - Crashman battle (although this might even stay in sync, since it depends on only his initial jump) - Quickman battle - Metalman stage after jester bots appear (lag) Parts not subject to (much) randomness AFAIK: - Woodman stage up until the apes - Woodman battle - 2nd half of Crashman stage after last substituted scroling - all of Quickman stage - Metalman stage up until jester bots appear Due to Murphy's Law all of the above are candidates for desyncing regardless. On the positive side, this game can be pretty reasonable on staying in sync when there isn't lag. Consider that the entire Heatman stage was successfully moved even though it was originally played after destroying Bubbleman.
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
CoolBumpty has explained it well. We had some brief discussion about this 8 pages back. Remember, the glitch in the Heatman battle is more than just for show.
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
re: Graveworm, I haven't found it any more difficult to do the ladder technique at that Woodman spot as compared to the Wily1 ladder. I use turbo and it usually takes many rerecords, but I do it without slowmode or frame advance. I watched the WIP. Great work guys! Dodging the energy drop in Crashman stage was a surprise. Only problem I noticed is that on the Quickman battle you didn't end on the highest platform (one of kolechovski's timesavers). Was there no way to do that without losing more than 2 frames in the fight?
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
Yeah, this is still missing from the Rockman tricks page. When I first tried to get it working, there weren't any .fcm files available so that I could watch the button presses. Now, at least you have Shinryuu's WIP for which you could load up in FCEU and frame-advance through the input display. Also, here is a short demo I made a while back that starts from a save-state: laddertech.fcm
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
gcc allows function nesting, as one of its many C extensions. C++ can't really do function nesting, except with a sort of hack. I'm sure Warp can explain why function nesting is generally a bad idea though. By the way, I think 4matsy and Highness best answered the editor question subject to the following constraint.
Truncated wrote:
under Windows
Emacs and Vim are likely not the most popular ones in this respect. Oh, and there's this about shooting yourself in the foot. edit: Just wanted to say sorry for continuing the off-topic discussion here. I guess a bit of zealotry is inevitable when it comes to talking about programming languages.
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
Mister Boat wrote:
http://s3plan.com/portman.mp4
Here are the results of my tests. Media Player Classic: works Windows Media Player 9: unrecognized file type; if forced, an error is detected then it plays without sound Windows Media Player 10: works Winamp 2: fails Naturally, deprecated media players will have more problems with newer formats, but at the very least this shows your graph is somewhat inaccurate.
Experienced Forum User, Player, Published Author (18)
Joined: 10/14/2005
Posts: 317
Hi flag. I did a few tests. ANSI C does allow you to use functions without declaration e.g. printf, but it doesn't specifically permit declaring global variables without a storage class. Consequently, although your first and third programs do compile under gcc, Intel C compiler 9.0, and Visual Studio .NET, they don't work with the IBM XL compiler. As for your second program, I don't see how you could use cout without first including <iostream>, and C++ generally isn't nearly as lax regarding those other things. Thus, I haven't gotten this program to build with any compiler. Edit: And as for output, I tried the top pi program across various compilers on four different operating systems and got the same output each time.
1 2
5 6 7
12 13