RoboZZle Forums

Some ideas for new game-play elements.

< Extension of the R/G/B color-scheme>

- Cyan, magenta, and yellow commands. Cyan will trigger on either blue or green tiles, effectively acting as a NOT-red condition; magenta functions as NOT-green; and yellow functions as not-blue.
~ Give new possibilities to players by giving them a NOT operator to use in their functions.

- Cyan, magenta, and yellow tiles and paint. Count as both of its component colors when the program checks to see whether it should execute a command.
~ Gives additional options for designers for puzzles and for ascetic qualities for their level.

- Gray and white tiles and paint. Gray tiles would only activate gray commands. White would activate any command, even those checking for a specific color.
~ Similar benefits to C/M/Y tiles.

- White commands. [Only useful if gray/white tiles are implemented.] Will always execute EXCEPT on gray tiles.
~ Similar benefits to C/M/Y commands.

< Miscellaneous >

- LOOP command. I'm thinking that these would work similar to the go-back command from Google's Turing Machine sketch. When activated they move the active position in the function back a set number of places. It would work similar to a recursive call, but wouldn't add to the stack and could go back to the middle of a function rather than the beginning.
~ Gives new options to players by allowing them to loop without calling sub routines and without creating stacks.

- IF command. Works similarly to the loop, only it jumps forwards a set number of places unless its condition is met. Allows for a collection of commands to be run based only on the criteria at the time of the first command without having to call a subroutine.
~ Allows players more flexibility when they can't spare subroutines to act as IF-THEN blocks.

- Multiple robots. [Based on denisb's parallelism idea.] Multiple robots placed at the start of the puzzle and all execute the same commands (although independently, so if one skips a command another may still follow it, and vise-versa). (Possibility of failure in the event of a collision?)
~ Allows puzzles that force players to clear two or more courses with only a single program. Although these types of puzzles may prove difficult to make and solve.

- Color-coded robots. [Only useful if multiple robots is implemented.] Allows players to color code commands (separably from tile-color coding). Only robots of matching color would execute the command.
~ Gives players an easier time with multiple robot levels by having parts of the code unique to each robot. Difficulty can be maintained via space constraints forcing robots to share code, or by making some/all robots in a level the same color.

- Wall. A tile that the robot can't move onto, it simply ignores the forward command if it would move it into a wall.
~ Functionally similar to pits, but without the instant failure condition. Good use may lead to interesting puzzles however. Could be used to interesting effect in multiple robot puzzles by allowing one robot to "stall" by repeatedly running into a wall.

- Ability to change profile name.
~ I seem to have forgotten to capitalize when signing up and would like to fix it. :P

- Large text-box for commenting on the forum.
~ It was hard to proof this due to only being able to see a small section of it at a time, I'm sure others feel similarly.

I was wondering about 2 new operators, non-deterministic ("?") and parallelism ("//").

The second one, parallelism ("//") could be added to any tile in function (like color) and would operate as follow: a second robot would be introduced on the board, at the current place of the first robot (like fork()), and would behave asynchron.
Example of questions (which should have to be answered)
Q: The number of robots on the board would be limited or not?
Q: What about join?
Q: One robot fall, the other die?
Q: Synchronous?
Q: How to count the step?
...

The first one, non-deterministic ("?") could be added to any tile in function (like color) and would operate as follow: when the tile would be execute, the robot would stack the context (board, position) before tile execution, and if the robot fall, the context would be bactrack, the tile skip the tile and continue with next tile.
Example of questions (which should have to be answered)
Q: How to cont the step?
Q: Should we do the current tile, at first, or when backtrack?
Q: "?//" in parallel one with current tile, one with next tile?
Q: Size of the stack of contexts, unlimited?

Best for you,
Denis B.

Puzzles that have -5 overall popularity or lower (calculated by subtracting the negative votes from the positive votes) are hidden in the puzzles list and do not count toward your puzzles solved on the scoreboard.
http://robozzle.com/forums/thread.aspx?id=186
Second comment from the bottom, igoro mentions this.

Thanks for your answer shaggy
But I made a test with a new level and a negative vote
And I got +1 in both scores
Anyway I never put 13 negatives votes before
So it is not the reason of this difference

@dorian_gray:
I think this is because of the fact puzzles with negative votes* are not counted in the official page with puzzles and the scoreboard. But when you open a player's page, it shows the number of all puzzles solved, even such with negative votes.

* > 5 negative compared to 0 positive, or more than 5 negative (do not know the exact function) if more than 0 positive...

Hello everybody
My first post
Nice game, thanks
I discovered it last month on my Windows Phone
I tried to create a new thread yesterday, it didn't work

About the count of solved puzzles,
Why is there a difference between the count in the scoreboard and the count in the player pages ?

IE:
In the scoreboard my count is 417 :
http://www.robozzle.com/scoreboard.aspx?from=100
And in my page the count is 430 :
http://www.robozzle.com/user.aspx?name=Dorian_Gray

???

Hello everybody
My first post
Nice game, thanks
I discovered it last month on my Windows Phone
I tried to create a new thread yesterday, it didn't work

About the count of solved puzzles,
Why is there a difference between the count in the scoreboard and the count in the player pages ?

IE:
In the scoreboard my count is 417 :
http://www.robozzle.com/scoreboard.aspx?from=100
And in my page the count is 430 :
http://www.robozzle.com/user.aspx?name=Dorian_Gray

???

Password recovery. :) I just can't imagine what it would be if I lost my password...

It would be interesting to see on puzzle page history of shortest solutions (length, date and recordsman nickname).

I would propose to allow players to submit puzzles only if they have 200 or even more solved puzzles.
Currently, some people solve the 40 easiest puzzles, have the rights to submit new ones and they post extremely useless puzzles!
I think seeing and thinking over more puzzles would minimalize submittion of such that are dumb. ;)

It'd be nice if there was better way to filter out all the dumb puzzles being submitted recently (although the problem is not new).

I'm wondering if it's possible to simply disallow any puzzle with a solution 4 commands or less. I'm 99.9% sure we've seen every possible interesting puzzle with 4 or fewer commands (probably even 5 or 6).

At first pass would be to not allow users to submit a puzzle where their own solution was 4 or less commands.

Long term a better solution would be to write a small server side solver that can check for any solution up to 4 commands. If there is concern about revealing "solver code", the solver could be bad with zero optimizations because 4 commands should still be near instant to do.

Not a feature, just usability idea. Please, place in the top of scoreboard history pages links to previous and next history pages (and actual scoreboard should be next to the last of history scoreboard). Also direct link to the actual scoreboard from any of scoreboard history pages should be useful.

To answer kraker's question, the reason it is on the iPhone version and not the Android or Silverlight version is because each is developed by a different person, so each has differing features.

And now that I understand your idea for displaying solution lengths in the actual puzzle screen, it does sound like a good feature. I always have to visit the puzzle pages to see what the shortest solution is and whether I have the shortest solution or not. That's the most important information on the puzzle pages to me, generally. It would also be nice to have a link near the Comments and Stats links that takes you to your solutions. That one would be trickier because as I understand it, the Silverlight client handles logins separate from the comments and solutions pages. In order to avoid cluttering the screen, the links and information could be viewed via a menu button that brings down the links and information when clicked and would be hidden otherwise.

Implementing the stack viewer sounds like it would be a rather large project while the other ideas seem much simpler to implement. Unfortunately, it seems development for Robozzle on all platforms has been stagnant for some time now. =/

Thanks Sticky. I have the Android version, found that first, then realised there was a browser based version as well. There's no stack display that I have found. No iThingies here to check that version.

I realise the stack gets huge. I mentioned a comfortable size, meaning maybe a 15 or 20 long window that could be scrolled over the entire stack. When you have the scroll bar at the top, the window follows new items being pushed or popped.

As for the shortest lengths request, at the risk of cluttering up the game screen, I figure an extra line or two under the existing Comments and Stats links is all it would take to save having to click through to other pages for the information.

Stack display, or partial stack display is a great feature! I've seen & admired it on the iPhone. It is really helpful in getting insight into Robozzle puzzles. Why isn't it available in the website Silverlight version?

I believe the iPhone and Android versions each display parts of the stack while solving. I think it would be quite difficult to show the whole stack at times since many puzzles can have a huge stack built up .

There is a page for each puzzle that shows both the shortest solution lengths by each player and your own shortest solution. If you're logged in, you can even view each of your solutions. These pages are accessible at the Puzzles link at the top of this page or while you're solving in the Silverlight version, you can click the "stats" link at the top left of each puzzle. There's no percentile, but you could always calculate that yourself if you want. Most puzzles have either few solvers or few solution lengths, so just looking at how you compare to others on the puzzle pages should be sufficient.

I am new here and am enjoying the puzzles immensely. Thanks for an interesting game.

Would it be possible to implement a stack display in the puzzle screen? It could grow with the stack up to a comfortable size, and then a scroll bar could allow the user to inspect different parts of it as the program runs or while paused. It might even be useful to allow 2 or 3 different areas to be viewed simultaneously. This would be a great debugging tool and instructional tool, and help users get a grasp of stacking operation and techniques.

Also, maybe a display on each puzzle screen of the shortest solution length, your shortest solution length, and a percentile of where you rank in all users solution lengths for that puzzle.

Don't want much, do I? ;)

Having solved all of the campaign puzzles, I think new ones should be added! As far as I can tell, all campaign puzzles were solved and nominated by igoro and it would be interesting to see what puzzles since the last puzzles were added would be nominated to be part of the campaign. Or if igoro isn't up to it, the community could nominate a few. This isn't really a feature request. I think the campaign is a decent compilation of good, high quality puzzles and could become a more extensive list of such puzzles.

Scoreboard history is interesting feature, thanks for it!
I've read that you can't reconstruct history of design scoreboard and as far as I understand there is similar problem with most shortest solutions scoreboard. I offer just to save current state at the end of each month and to show at least “robozzle new age” history:)
(I know this is crutch, but it is easy to implement and it will work:))

I had a fun idea the other day about the stars we collect in robozzle.

I think it would be awesome if there was a new star collecting system introduced into robozzle where there are different colored stars and these stars need to be collected in order. (for example, yellow star first, white second, black last).

This could be fun. It was just an idea. Take it or leave it. Could be a lot of work for the programer/s aswell. haha.