Replying to Truncated's comments...
> > using Truncated's idea (route Q), the first zipping is left rather short> > (the buildup takes too much time) - cost-effeciency for the 4 magnet beams> > is bad> > You could use only the second part, if it's faster than recharging a second> time. But I doubt it, see below.
Indeed...
However, using the route P has another good feature: it avoids
the red jumping junks' lag completely.
> > route A has quite much delay in it> > Why not switch to F directly when you fall down from the beam at 23320,> shoot the cannon to the right and pick up the refill as you walk past> it at 23410? Is it because the magnet beam is still on screen?
Yes, weapons can't be switched as long as there are weapon artifacts
on the screen. (Curiously, it doesn't apply to the select-button pause.)
Admittedly I didn't switch as soon as possible. I switched immediately
after picking the first powerup. I didn't think of the method you suggested.
The beam to the hole has to be made between the powerups though.
> Also you seem to hit the corner at 23313. Turn backwards to just miss the> corner or incorporate the switch to F-weapon in the fall, since you only> need the vertical motion anyway. (Again, perhaps not possible since the> magnet beam is still on screen.)
Hitting the corner is not fatal, because there is no horizontal acceleration
delay in jumping. But the weaponswitch can't be done yet.
However, I purposefully delayed the magnet beam as late as possible.
If I had released the beam 1 frame earlier, the endpoint of the zipping would
have been 16 pixels earlier (avoiding the corner). This would allow for switching
weapons sooner. But, Rockman would drop onto the cannon, and the delay caused by
that hit is not avoidable by pressing pause.
> Are you already at max falling speed since you're walking off a magnet beam?
No, since Rockman technically wasn't walking on the magnet beam.
He was zipping inside a wall.
Rockman experienced only a normal dropping (from ceiling), no magnet beam.
> If all this is impossible for some reason, why not just recharge twice at> the cannons + flying bombs later? You have enough ammo to last until then.> Two extra weapon switches, but it looks faster than what you are doing in> route A at the moment.
I initially rejected it precisely because of the weapon switches,
but you may be right.
It's possible to switch weapons 3 frames before the first cannon.
> - 26030, why don't you go trough the wall here as your plan suggested> (route N)? Is it undoable?
You mean route K? (N there stands for not doing it)
I didn't do it because there wouldn't have been enough room to construct
a full-length beam after that.
The extra beam in the C/D room saves about 30 frames, by the way.
Beam S (after Bombman's door) would have saved about 16 frames.
> In Bombmans lair, why don't you miss the corner as close as possible when> going down? Is it because you will hit the enemies? If so, how about taking> them out with the F-shield?
Yes, I would have hit the enemies. Also, F-shield is out of question
because of the immense lag it creates in that tunnel.
However, I should try to take the hit (before the door) as late as
possible in order to pass through the first eye blob with the blinking.
The only place where I can see that magnet beam could still be very useful,
is at B and H. H is out of question because it requires 7-8 extra beams.
B could be considered after all.
The gain of F is rather small (and it's very difficult timingwise,
because the beam must not be too long!)
Using Q-aY-B, there would be 11 beams used (3 free) before the first ladder
- not enough left for the 4 beams in the beginning of the next long road.
That doesn't work.
P-B would have 13 used before recharges - taking two recharges, there would
be 3 used (11 free).
Route E eats 3. -> 8 free
Route C eats 2. -> 6 free
Using one extra in C/D room -> 5 free
Route N eats 3. -> 2 free
The 1up room eats 2 -> 0 free.
This route sacrifices aY and F for B.
Do you think this is this worth testing?
Joined: 7/17/2004
Posts: 985
Location: The FLOATING CASTLE
I like bombman overall, but I agree that the part where you go back to get the powerup from the multishot ground gun looks slow. What if you switch to elec or fireball when approaching it, blast it early, then switch back to M. You'd have just as many switches with no backtracking. Or is it not possible to manipulate the powerup then?
Using Q-aY-B, there would be 11 beams used (3 free) before the first ladder
- not enough left for the 4 beams in the beginning of the next long road.
That doesn't work.
P-B would have 13 used before recharges - taking two recharges, there would
be 3 used (11 free).
Route E eats 3. -> 8 free
Route C eats 2. -> 6 free
Using one extra in C/D room -> 5 free
Route N eats 3. -> 2 free
The 1up room eats 2 -> 0 free.
P-B is still untested.
Actually, the one "extra" should probably be used to start the B path earlier.
Instead, I tried to make bY (getting the second powerup in the bottom right corner) faster, but there are the following problems:
- If I do maximum length zipping, the weapon selection box will become available only a moment before hitting the first recharge item. Therefore, the previous way was actually the fastest, assuming the full-length zipping. (Zipping can't be started earlier, because it depends on the existence of floor at room bottom.)
- If I do a shorter zipping, the weapon selection becomes available earlier, but so far I haven't been able to luck-manipulate the big recharge item in that location again... (I tried both the fire weapon and the basic weapon.)
And you've been trying everything that might make the powerup appear, like hitting Up and Down every other frame, or holding Select for a while?
Hitting up/down has no effect to the randomness unless it somehow affects the progressing speed (by grabbing a ladder).
Of course I haven't tried everything. There are quite a number of different combinations of actions you can do in ~20 frames.
I wanted to try to give some more input on this, but my connection at home is fried for some reason, so I can't. :( I am currently at my lab, but I can't watch the movie here. Sorry.
Finalfighter's idea didn't work, but I ended up improving the level anyway.
The Elecman level is now 73 frames faster due to a most basic optimization.
Also, Fireman level is now 1 frame faster.
Finalfighter has also discussed a new idea regarding the Bombman level. I'll try it out before revealing what it is.
In any case, the WIP is now updated.
https://files.tasvideos.org/bisqwit/bisqwit-rockmanv6-incomplete.zip
Oh, and I've added lots of animations at the RockmanTricks page.
RockmanTricks
Wow, I like it, magnet beam glory.
I just had idea, in Fireman level, when you arrive near Fireman room, you had to climb down ladder zigzaging. Would shooting magnet beam on wall zip you pass those wall? (see pic below)
*edit, the "Special case: Walls at the top of the screen" trick, no good?*
its been more than a week, i want to know if you made any progress in it?
Since the last update of the WIP, the following happened:
- I started finally making the GIF files for the RockmanTricks page.
- To ease the making of those animations, I created a level editor so I could create custom maps for better demonstrations.
- I started envisioning creating a patched version of Mega Man that has entirely new levels. I'm still working on that. :)
I haven't yet played the movie further. (That is, after the second strip of the Bombman level).
------
If someone wants to have the sourcecode of the level editor, grab it here:
http://bisqwit.iki.fi/src/arch/rockedit-0.8.1.tar.bz2
"decode" reads the "megaman1.nes" file and outputs the sourcecode that may be put to the various header files (wily*.hh, *man.hh). The source code contains the tables that define the layout of each level.
"encode" reads "megaman1.nes" and writes "tmp.nes" with new levels. Editing the levels requires always a recompilation of the "encode" program.
This has probably been brought up before. Nevertheless, I was wondering, prior to executing the magnet beam zipping glitch, what's the reason for jumping to place the magnet beam then jumping again to get on it, as opposed to jumping once to place and then get on it?
"Great spirits have always found violent opposition from mediocrities. The latter cannot understand it when a man does not thoughtlessly submit to hereditary prejudices but honestly and courageously uses his intelligence."
This has probably been brought up before. Nevertheless, I was wondering, prior to executing the magnet beam zipping glitch, what's the reason for jumping to place the magnet beam then jumping again to get on it, as opposed to jumping once to place and then get on it?
The beam has to placed at exactly the right height.
When you press A (jump), Mega Man starts elevating at maximum speed (about 4 pixels per frame).
Due to the time quantization (the game is animated once per each frame), he might never hit the exactly right height, because one frame he's below and next frame he's already above.
In order to get the right height, you need to release A before releasing the beam (when you release A, the speed decreases and thus the precision increases), and when you do that, you can't re-elevate without jumping again.
In some places, it's possible to do it in 1 jump (when the floor is at certain height). I haven't mapped the exact circumstances. It seems to also depend on the frame of animation Mega Man is walking in (affects the height by 1 pixel).
Alright, I was thinking it was something like that. So techinally isn't there a 1/4 chance you could do it in one jump, that is, rockman will be at the right pixel height during his 4 pixel/frame velocity for one of the frames? Then you could also adjust things like a previous magment beam he'll be jumping from by placing it acurately with the 2 jump method, and then this walking animation 1 pixel adjustment as well. Are you trying to do this, because it seems like you use the 2 jump method much more than 3/4 of the time.
"Great spirits have always found violent opposition from mediocrities. The latter cannot understand it when a man does not thoughtlessly submit to hereditary prejudices but honestly and courageously uses his intelligence."
Alright, I was thinking it was something like that. So techinally isn't there a 1/4 chance you could do it in one jump, that is, rockman will be at the right pixel height during his 4 pixel/frame velocity for one of the frames?
Yes, something like that, but it's not timing/randomness related, except for the vertical 1 pixel variation in walking animation.
Spider-Waffle wrote:
Then you could also adjust things like a previous magment beam he'll be jumping from by placing it acurately with the 2 jump method, and then this walking animation 1 pixel adjustment as well. Are you trying to do this
Yes, I'm trying it in some places. (Only when I need to, actually...)
Usually I just try to maximize the supporting beam heights so that the number of required beams is minimal.
*edit, the "Special case: Walls at the top of the screen" trick, no good?*
