The Road to Tile Factory 2

[Jonathon Duerig]

I've gone through all of the suggestions people have made for the game
and thought up a few things myself. Below is the current todo list for
the release of Tile Factory 2. I also have a 'Maybe' list of things
that I haven't decided whether I want to do or haven't figured out how
to do properly. Feel free to post any more suggestions here. I'll keep
you posted on my progress:

----
TODO
----

New Maps (Need ~40)
Drag Scrolling, hotkeys for scrolling
Stamps
Locks (Red Zones)
Barriers impermiable inside locks
Always-on Play tab
Add a 'Super-fast-forward' button
Hotkeys for tabs/parts
Hotkey for play/stop
Challenge badges
Display time/part/item/breakage during play
Integrate with a level management system
Distinguish between solutions and levels. Dual-load all the time.
Load levels/solutions using URLs
Select square then move/delete
New Parts: Breakpoint (pauses simulation when powered), Lever (click
to toggle power while playing), Light Bulb (whole-tile power
indicator), Random (on 50% of time when input is powered, off
otherwise), Label (reveals custom text on mouseover)
Show current wires on parts when mousing over or moving them even when
not in wires mode
Display wire count
Translations to Spanish/French/German. Other languages?
Allow larger window sizes
Place any color paint
Simulation continues even with victory screen up
Track display issues
Ungroup disconnected tiles when using solvent.

-----
MAYBE
-----

Story? Ideas?
Allow loading of recent levels?
Leaderboards?
Limit number of pieces of each type?
Rotation of entire mosaics? Problematic collision detection,
animation.
Create reversed wires?
Comparison Part? A kind of sensor which turns on when all adjacent
comparison parts have the same item on top of them.
Can place items on top of other parts?
Configuration library?
Color-blind support?

[SaintPeter]

Oh, yeah, I'm impossibly excited about this new version!

Some great new features that should make for some awesome puzzles. I
can't wait to see what you and my fellow Tile Groupies come up with.
The mind boggles!

[SaintPeter]

Will you be releasing an "official" 1.5 with the new wire mode and
interim fixes? I've been using the "wire-test" to build my solutions
because it is almost impossible to build complex logic with the 1.4
wire interface. I live in fear that you will delete it and I'll be
stuck.

Maybe you should divide the above features into "New Features" and
"Enhancements" and roll out the "easy" Enhancements first. Then you
can work on the entirety of the new 2.0.

Simple Enhancements/Fixes might be:
Hotkeys
Wire Display Changes
Super-Fast-Forward mode
Track Display

Of course . . I'm not writing it, so it's easy for me to say. I'm
just hoping for a stable version to fiddle with until v2.0 comes out.

RE: Lever
Doesn't that completely change the game? All logic becomes simple
when you have a human brain doing the driving. Unless it's something
I don't understand, it sounds like a game breaker.

[Jonathon Duerig]

I have pushed 1.41 to Kongregate which includes the wiring changes
(should be even better than wire-test) and a couple of minor bugfixes.
I don't plan to do any more releases until 2.0.

Regarding the lever, on most puzzle levels it won't be available for
use or may be pre-placed and wired just as the copier is usually
restricted. I think where the lever, the light bulb, and the label
will be most useful will be in player-made toys.

-D

[barryd15]

@SaintPeter:
Your comments re. enhancements vs. features seem to make a lot of
sense.

Re: Lever:
I suspect the idea is that lever will not be available for most
puzzles. Remember that the puzzle designer gets to choose which tools
are available. Although a feature may be potentially game-breaking,
it's safe to include it so long as the designer can prohibit it for
their puzzle.
Makes me wonder if the part buttons should no longer have fixed
positions, but rather the positions should be assigned based on which
parts are available for the puzzle. That might make it more practical
to have a large number of parts in the game (think lemmings 2).
Hotkeys which *don't* change would still be nice...

[Goldbear]

What I totally miss is the ability to turn off the background music
alone.

Sometimes it bugs me so I turn it off but then I miss the clicks and
crashes.

[Jonathon Duerig]

Goldbear, I'll see if I can add this. I just didn't want to clutter up
the interface too much.

-D

[geironul]

This is not my original idea, but I like it so i will repost. (I read
it somewhere on here but couldn't find it again). Having the ability
to place wires in series or more than one to the same block would be
really useful. I don't know if you saw this suggestion and decided it
wasn't possible, or just missed it. It would probably help if you put
a list of things you considered and decided not to include so you
don't get a bunch of repeat suggestions.

[barryd15]

I sometimes use the "Rate this post" feature to indicate suggestions
that I agree with, without making another post. Of course it's a bit
tricky if someone suggested several features in the same post, but I
often agree with all of them, so it works out.

[Goldbear]

On 17 Sep., 18:57, Jonathon Duerig <tyrec...@gmail.com> wrote:

> On Fri, Sep 17, 2010 at 10:48 AM, Goldbear <mhabluet...@gmail.com> wrote:
> > What I totally miss is the ability to turn off the background music
> > alone.
>
> > Sometimes it bugs me so I turn it off but then I miss the clicks and
> > crashes.
>
> Goldbear, I'll see if I can add this. I just didn't want to clutter up
> the interface too much.
>
> -D

Personally, I'd feel satisfied, if the TOGGLE button in the MENU was
splittet, one side for music and the other for sounds. But I think a
slick note above or below the speaker in the interface wouldn't hurt :)

[Neebat]

Something like holding CTRL instead of SHIFT so that when you click to
end a wire, it automatically starts a new wire from the previous
wire's destination. Is that what you mean?

[Neebat]

When I don't want sound (at work), I turn off the sound in the OS. I
don't need a sound toggle at all.

When I want sound without music, (most of the time), I need the game
to support it. So I wouldn't mind at all if the current sound toggle
were replaced with a music toggle. Might upset some people though.

[Jonathon Duerig]

On Fri, Sep 17, 2010 at 12:40 PM, Neebat <newel...@gmail.com> wrote:
> Something like holding CTRL instead of SHIFT so that when you click to
> end a wire, it automatically starts a new wire from the previous
> wire's destination.  Is that what you mean?

geironul, Neebat, this feature is currently under consideration
(listed as 'Create reversed wires' under MAYBE). I've not come to a
firm conclusion about the best interface and implementation for it.

-D

[Neebat]

In the interest of helping you prioritize...

> Drag Scrolling, hotkeys for scrolling

YAY! Hate those vertical scroll arrows. I WISH we could see this
before 2.

> Locks (Red Zones)

If you just forbid moving things into red zones, level designers
wouldn't need so many stupid barriers.

> Barriers impermiable inside locks

I'd rather see barriers being ALWAYS impermiable. All you need is a
way to set the length of a conveyor shorter.

I really hate using barriers to end conveyors. I don't like having to
choose between "optimal" and "good-looking" just because the optimal
solution has conveyors running off all the edges of the screen.

> Always-on Play tab
> Add a 'Super-fast-forward' button
> Hotkeys for tabs/parts
> Hotkey for play/stop

> Display time/part/item/breakage during play
> Load levels/solutions using URLs
> Allow larger window sizes
> Display wire count

YAY! These are all AWESOME.

> Challenge badges

> Integrate with a level management system
> Distinguish between solutions and levels. Dual-load all the time.

> Stamps

> Translations to Spanish/French/German. Other languages?

> Place any color paint
> Simulation continues even with victory screen up

> Configuration library?
> Color-blind support?

Meh. I really don't care about these.

> Select square then move/delete

What about copy/paste? If you could copy a square to the clipboard
and paste it into another browser tab, that would be awesome!

> New Parts: Breakpoint (pauses simulation when powered),

Please do not do this. Breakpoint should be a setting on a part, not
a whole separate part. Maybe CTRL-click a part toggle whether acts as
a breakpoint?

> Lever (click
> to toggle power while playing), Light Bulb (whole-tile power
> indicator), Random (on 50% of time when input is powered, off
> otherwise), Label (reveals custom text on mouseover)

Sounds nice for "toy" levels. They should be off-by-default in the
Buttons tab.

> Show current wires on parts when mousing over or moving them even when
> not in wires mode

Even when dragging a part, pretty please!

> Track display issues
> Ungroup disconnected tiles when using solvent.

I don't really understand these. Could you elaborate on the second
one, please?

> Limit number of pieces of each type?
> Rotation of entire mosaics? Problematic collision detection,
> animation.
> Create reversed wires?

> Can place items on top of other parts?

These all sound good. Definitely not as important as the others.
Don't do collision detection DURING the rotation. Only the final
position.

> Comparison Part? A kind of sensor which turns on when all adjacent
> comparison parts have the same item on top of them.

Interesting, but not important. I'd much rather see the "UPDATE" part
I wrote about in another thread.

[barryd15]

Neebat, I believe what you described has previously been referred to
as "wires in series".

Reversed wires - holding CTRL while clicking the endpoint of a wire
will result in the wire going in the opposite direction. By itself,
it's not too useful, but in combination with SHIFT, it effectively
allows you to quickly specify many wires with a common *destination*,
just as SHIFT normally allows you to quickly specify many wires with a
common *source*.
Wires in series - was also suggested, but no control scheme for it
was suggested, and I think it's lower priority because simply clicking
twice on each intermediate cell will accomplish this without too much
trouble.

I think Jonathon already understood the proposal and the benefit of
reversed wires, but after seeing Neebat's post, I just wanted to
clarify the 2 separate suggestions.

[Neebat]

I don't think reverse wires and chained wires are the same thing.

Valid destinations: logic, set, reset, mem, light-bulb, conveyor,
sprayer, mixer, copier
Valid sources: logic, mem, switch, sensor.
Valid chained parts: logic, mem

Reverse wires:
1. While holding a meta-key, click a valid destination. The wire
attached to the cursor should point to that destination.
2. Click a valid source. It makes a wire from source to the
destination. And the destination stays selected.
3. Repeat 2 and all wires go to the same destination.

Chained wires:
1. Click a valid source.
2. While holding a meta-key, click a valid chained part. It makes a
wire from source to chained part. AND it selects the chained part as
a new source.
3. Repeat 2 and you get a chain of wires leading from A to B to C to
D.

Those both seem like VERY good ideas. You could even combine the two
to draw a chain backwards.

On Sep 17, 1:44 pm, Jonathon Duerig <tyrec...@gmail.com> wrote:

[Jonathon Duerig]

On Fri, Sep 17, 2010 at 1:18 PM, Neebat <newel...@gmail.com> wrote:
> I don't think reverse wires and chained wires are the same thing.

I am considering doing reverse wires. I don't think I will do chained
wires. The reason is that, as barryd states, chained wires can already
be done pretty easily. Click source, double click dest, double click
next dest, etc. Implementing a special key to do chaining simply turns
a double-click to a click. By contrast, reverse wires may save much
more work relatively speaking.

-D

[Neebat]

I agree. I was just interpreting geironul's request, which sounded
like chaining to me. And it sounds like a good enough idea to be a
"nice-to-have"

Note, reverse wires can use the same hotkey as repeated wires from the
same source. To reverse a wire, the meta key (whether it's SHIFT,
CONTROL or something else) has to be active when you select the first
part, but doesn't have to be active after that. Chaining and
repeating only care whether you hold the hotkey for the second (and
subsequent) parts, so they need to be different hotkeys.

I have a minor request in relation to this. It would be awesome, if
I've just used shift to place 3 parts or 3 wires, and I let up on
shift, if it would delete the part or wire currently on the cursor.
As it is, I usually don't let up on shift and end up with a part I
need to junk.

On Sep 17, 2:23 pm, Jonathon Duerig <tyrec...@gmail.com> wrote:

[Iván Nieto]

On Sep 17, 9:31 am, Jonathon Duerig <tyrec...@gmail.com> wrote:
> Locks (Red Zones)

How will these work? I can see three things this may mean:
- can't place/remove wires.
- can't power/depower initial copiers/others.
- can't move tile/items in/out the locked zone.

> Barriers impermiable inside locks

I see this as a "wall", similar to the ones surrounding the factory.

> Allow larger window sizes

Another option is to reduce the tile's size. Although difficult to do,
maybe a "small tiles" option?

> Challenge badges

> Limit number of pieces of each type

These two could be put to work together. Add challenge modes to a
puzzle, that just limit the number of pieces, or time allowed.
Now, when the number of pieces is limited, I'd think it would be a
good idea to allow surpassing that limit, but not accept the solution.
When optimizing, normally you start with a non-optimizing solution
then slowly remove pieces. Having a hard limit would force you to
optimize from the get-go, and that is much more difficult. We don't
want people editing the map just to remove the limits so they can find
a solution, then painstakingly copying the solution to the original
map.

> Place any color paint

I'm guessing this is any of the 16 colors, not new, custom colors (I
notice a severe lack of orange, for example.)

> Configuration library

What's this?

> Select square then move/delete

Please, combine this with a shift-click to add/remove parts to the
selection.

> New Parts: Breakpoint (pauses simulation when powered),

I don't like this. Even if the part is not counted at the end,
sometimes it would need sensors/none/alls for it to function, and
those would count unless removed. I'd rather have instant-fastforward
(no animation at all) to a specific turn, with next/previous 1/5/20
turns.

> Lever (click to toggle power while playing)

This would "action", which I oppose to in a game like this, but would
be extremely useful while testing as long as they are not allowed at
the final solution.

> Light Bulb (whole-tile power indicator)

Just an indicator? A prettier thing to use on those N-bit counters,
for example?

> Random (on 50% of time when input is powered, off otherwise)

Ugh, I don't like randomness either, although I can see the potential
for self-adjusting machines. Will be a pain for the "fastest record",
though.

> Label (reveals custom text on mouseover)

Useful for helps and tips, nice. Make sure they don't count for
scoring ;-)
Unless they can also work as barriers...

> Show current wires on parts when mousing over or moving them even when
> not in wires mode

I'd also ask, maybe in wire mode only, when hovering over a piece, to
highlight the pieces that are wired to it.
Sometimes a piece is wired to many other pieces clumped together and
it's difficult to see which ones.
This is specially hard to see in test mode when a wire comes from a
sensor in two parallel conveyor belts covered with wires. You can't
see where the wire starts. Maybe show wires at the topmost layer?

> Comparison Part? A kind of sensor which turns on when all adjacent
> comparison parts have the same item on top of them.

I'm not sure I understand the concept, but seems interesting. How
about this: a Comparison Part that can only receive wires from
sensors, and only powers up if all sensors have the same thing over
each of them.


On Sep 17, 3:02 pm, Neebat <newell1r...@gmail.com> wrote:
> > ...

> > Ungroup disconnected tiles when using solvent.
> I don't really understand these. Could you elaborate on the second
> one, please?

eNoNw4EJADAIA7BqRUdh+v+1m4EYoMJJ85axB9T12L4pJaeLAhCY9wEaAAIU

> > Rotation of entire mosaics? Problematic collision detection.

> Don't do collision detection DURING the rotation. Only the final
> position.

Even on a very long and thin mosaic that swipes over the whole factory
when rotated?

[Neebat]

On Sep 17, 3:38 pm, Iván Nieto <ivan.niet...@gmail.com> wrote:
> On Sep 17, 9:31 am, Jonathon Duerig <tyrec...@gmail.com> wrote:
>
> > Allow larger window sizes
>
> Another option is to reduce the tile's size. Although difficult to do,
> maybe a "small tiles" option?

I asked about that. He said the tiles would be unrecognizable if
shrunk. Certainly the wire-cloud would become even more dense.

> > Challenge badges
> > Limit number of pieces of each type
>
> These two could be put to work together. Add challenge modes to a
> puzzle, that just limit the number of pieces, or time allowed.
> Now, when the number of pieces is limited, I'd think it would be a
> good idea to allow surpassing that limit, but not accept the solution.
> When optimizing, normally you start with a non-optimizing solution
> then slowly remove pieces. Having a hard limit would force you to
> optimize from the get-go, and that is much more difficult. We don't
> want people editing the map just to remove the limits so they can find
> a solution, then painstakingly copying the solution to the original
> map.

Anyone ever played World of Goo? The "OCD" bonuses are like that.
You play the level as normal, but if your complete solution meets
certain conditions, you get an added bonus. There could be different
types of OCD conditions, like don't use more than N memories or finish
in under 20 seconds, or never turn on copier X.

> > Configuration library
>
> What's this?

I wondered that too.

> > Select square then move/delete
>
> Please, combine this with a shift-click to add/remove parts to the
> selection.

That would be cool. Not a huge thing, but nice to have.

> > New Parts: Breakpoint (pauses simulation when powered),
>
> I don't like this. Even if the part is not counted at the end,
> sometimes it would need sensors/none/alls for it to function, and
> those would count unless removed. I'd rather have instant-fastforward
> (no animation at all) to a specific turn, with next/previous 1/5/20
> turns.

That's one reason I say the breakpoint should be a flag on any part,
not a new part.

> > Lever (click to toggle power while playing)

> > Light Bulb (whole-tile power indicator)

> > Random (on 50% of time when input is powered, off otherwise)
>

> This would "action", which I oppose to in a game like this, but would
> be extremely useful while testing as long as they are not allowed at
> the final solution.

I think the idea here is to make toys where you can flip a lever to
make something happen while a level is running. They're definitely
not something you'd use in a normal level.

> > Comparison Part? A kind of sensor which turns on when all adjacent
> > comparison parts have the same item on top of them.
>
> I'm not sure I understand the concept, but seems interesting. How
> about this: a Comparison Part that can only receive wires from
> sensors, and only powers up if all sensors have the same thing over
> each of them.

Think of it as a goal tile. You put a tile on each comparator and if
they match, they light up. I could see it being useful as a goal
condition.

> > > Ungroup disconnected tiles when using solvent.
> > I don't really understand these.  Could you elaborate on the second
> > one, please?
>
> eNoNw4EJADAIA7BqRUdh+v+1m4EYoMJJ85axB9T12L4pJaeLAhCY9wEaAAIU

Ok, now THAT I understand. It is a funny effect. But I've never seen
a reason to unglue tiles.

> > > Rotation of entire mosaics? Problematic collision detection.
> > Don't do collision detection DURING the rotation.  Only the final
> > position.
>
> Even on a very long and thin mosaic that swipes over the whole factory
> when rotated?

Absolutely. You used the keyword there, "over". We can assume the
rotator lifts the mosaic to turn it, and thus it clears everything
until it's finished rotating.

[geironul]

> geironul, Neebat, this feature is currently under consideration
> (listed as 'Create reversed wires' under MAYBE). I've not come to a
> firm conclusion about the best interface and implementation for it.

OK, I just liked the idea and wasn't sure if 'create reversed wires'
included the multiple to one object or if you just meant that you
could click the destination then the source for a wire. I also agree
that daisy-chain is not that important, since it only removes one
click. I was just trying to remember everything that was in the
original post (which I still haven't found).

[CryoGenik]

I cannot freaking wait to get my hands on the second version of this
game. All these ideas sound awesome so far. I have a few quarks here
and there but they are pretty much the same concerns as the general
concensus. Besides I trust in John's judgement and predict that
whatever he puts in or out the next version is going to be awesome :D

[Lynlially]

For the music toggle thing, I suggest putting it in the right-click
menu ?
Screen space wouldn't be a bother anymore.

As for the translation to French, if you don't have someone yet, I
could do it,
though I don't really think you'd need a translator if you're up
yourself to spend some time on google translate,
since the stuff to translate are mostly single-words.

And I really like how most of my suggestions end up in the final todo
list ^^

[Foreman]

In additional to the badges of challenge, I hope that there could be
some more contitional badges for each map, just like fantastic
contraption 2:

Flawless: no breakages
Luddite: no logic parts are used
Streaming: no MEM parts are used
FlatField: no barrier parts are used

AlwaysOn: no conveyer parts are being powered-off
Consistency: all initial items at source edge of copier are neither
moved or broken
AdhesiveFree: no unused glued tiles on final product

[Neebat]

No breakages should be the easiest challenge.

[SaintPeter]

RE: Challenges
It might be interesting to have a bronze/gold/silver for time,
although I'm not sure how you would determine the edge points for
those. If this group has demonstrated anything, it's that there are
almost as many ways to solve these challenges as there are players.

[geironul]

To me the simple ranking by time, parts, breakage and items can get a
bit confusing. Here are some thoughts on how to tie them together:

Another way to score levels
=============================
It seems that you could assign a monetary value to each part and item
in the game, then add all the prices up at completion and have one
value for each level. The lower the price, the better. It would give
us a single score to aim for, and allow an easy way to create badges
or gold/silver/bronze.

You could either make all 'items' one price, make all 'parts' a
different price and multiply by the number of each to get a total, or
you could price each 'item' and 'part' individually. I imagine doing
individual prices would be rather difficult to balance properly.

This would also allow you to differentiate between 'start-up' cost and
'running' cost. Start up cost would be the price of all the parts,
'running' cost would be the price of all the items.

Map Completion
==============================
If you implemented the pricing as above, you could set goals on each
map. A time of course, a 'start-up' cost and a 'running' cost. You
could make gold/silver/bronze completions. I like the fact that you
can build a complex monstrosity (I usually do) and complete a level.
So I don't think exceeding the goals should stop completion of a map,
but maybe you can unlock more levels if you complete all the maps on
an island with bronze. You could do this without the pricing just by
setting 'part' and 'item' goals, but I think pricing might be more
intuitive and allow a bit more wiggle room in individual designs.

If you priced logic parts separate from the other parts, Luddite
solutions now mean a logic parts price of $0. Engineer solutions would
have the lowest 'running' cost. Optimal size could be achieved by
setting a price for one square of factory space, the lower the price
the better.

I think the pricing could be incorporated into a storyline pretty
easily too.


On Sep 17, 7:31 am, Jonathon Duerig <tyrec...@gmail.com> wrote:
> I've gone through all of the suggestions people have made for the game
> and thought up a few things myself. Below is the current todo list for
> the release of Tile Factory 2. I also have a 'Maybe' list of things
> that I haven't decided whether I want to do or haven't figured out how
> to do properly. Feel free to post any more suggestions here. I'll keep
> you posted on my progress:
>
> ----
> TODO
> ----
>
> New Maps (Need ~40)

> Drag Scrolling, hotkeys for scrolling

> Stamps
> Locks (Red Zones)
> Barriers impermiable inside locks

> Always-on Play tab
> Add a 'Super-fast-forward' button
> Hotkeys for tabs/parts
> Hotkey for play/stop

> Challenge badges
> Display time/part/item/breakage during play

> Integrate with a level management system
> Distinguish between solutions and levels. Dual-load all the time.

> Load levels/solutions using URLs
> Select square then move/delete
> New Parts: Breakpoint (pauses simulation when powered), Lever (click

> to toggle power while playing), Light Bulb (whole-tile power

> indicator), Random (on 50% of time when input is powered, off
> otherwise), Label (reveals custom text on mouseover)

> Show current wires on parts when mousing over or moving them even when
> not in wires mode

> Display wire count

> Translations to Spanish/French/German. Other languages?

> Allow larger window sizes

> Place any color paint
> Simulation continues even with victory screen up

> Track display issues

> Ungroup disconnected tiles when using solvent.
>

> -----
> MAYBE
> -----
>
> Story? Ideas?
> Allow loading of recent levels?
> Leaderboards?

> Limit number of pieces of each type?
> Rotation of entire mosaics? Problematic collision detection,
> animation.
> Create reversed wires?

> Comparison Part? A kind of sensor which turns on when all adjacent
> comparison parts have the same item on top of them.

> Can place items on top of other parts?

> Configuration library?
> Color-blind support?

[barryd15]

Re: overall solution score

I like the idea of calculating an overall score by weighted
combination of multiple metrics.

Eliminating a couple of parts is good, but not if the production run
takes 30 seconds longer. And saving a couple of seconds is good, if
it doesn't require 30 extra parts.

When I'm scanning people's solutions looking for an all-round good
("balanced") solution, I'll start with the ones with a good rating for
this overall score. The weightings might be a little arbitrary, but
hopefully they won't differ too much from what most people might
expect. When in doubt, Jonathon can use the idea of money and profit
as a guideline.

I like the idea of using "money" to make the combined score more
concrete and easy to grasp, but with our current model of producing a
fixed number of products, it may be difficult to incorporate runtime
into a "money" score in a sensible way.

[barryd15]

(This post is made of stuff I chopped from my last post because it was
too half-baked.)

Re: per-level awards

A couple of award types have been suggested:
* bronze/silver/gold - differ in quantity
* luddite/engineer - differ in quality, worked well in Crayon Physics
Deluxe

I realize they're not incompatible with each other, but we probably
don't want too many different awards which would get kinda
complicated. I thought we should collapse bronze/silver/gold into a
single award based on achieving a target value for an overall solution
score, which would fit nicely alongside luddite and engineer.

Problem: The target value needs to be set for each level, and at
publishing time, the level author will likely only have a single
primitive solution on which to base their decision.

Really half-baked ideas:
* call the combined rating "cost" and call the award "scrooge" (for
minimising cost) (as mentioned in previous post, difficult to
incorporate time in this)
* combined rating is "profit" and award is "entrepreneur", for turning
a positive profit after a fixed number of seconds
* but then instead of having to set a target value for the level,
you have to specify the sale price for each product and/or the
available time in which to turn a profit

This leads to the following ideas:
The user solves the puzzle and starts the simulation. The simulation
keeps running until the user tells it to stop. Different awards are
granted for the level as the conditions for each award are met. It is
the user's responsibility to have expectations of which awards their
machine will achieve and stop it once they are all received. The
conditions for "passing" the level in TF1 (ie. produce 5 of each goal
product) take the form of just another award (eg. "low-volume
producer") with no special treatment in TF2.
In sandbox spirit, conditions for progression to later levels, if any,
should be light, eg. earn any award from each of the prerequisite
levels.
Levels which receive an award are automatically filed away in the
stored flash data in such a way that the author can easily say eg. "I
want to see my luddite solution/s for this level" (very half-baked
idea here).

Anyway, sorry for putting out all these conflicting and half-baked
ideas...

[barryd15]

I just thought I should mention the existence of a game called
"Silicon Foundry" by the same guy who did Codex (of Alchemical
Engineering) and Manufactoid. Like Manufactoid, it's non-browser and
so more obscure than Codex. I haven't really played it, but well, it
incorporates money and profit into an engineering game so I thought
people might find it interesting. Anyway, I acknowledge that TF2 is
intended to be a very very different game to that. But, it's
interesting nonetheless.

Hope I'm not upsetting anyone by mentioning another person's game
here.

[SaintPeter]

I played Silicon Foundry (Love the Zachtronics games, which Johnathn
credits as inspiration) and the scoring system was pretty crap. It
wasn't balanced at all, although it may have been a realistic
depiction of silicon economics.

[Jonathon Duerig]

I've been working a bit on implementing stamps and this has made me
realize that I can do extra stencils more easily than I had previously
thought by moving to raster rather than vector representations. So the
stamp designs people have posted will actually become new stencils in
the game. In addition, people will be able to create custom stencils
for a level. This should make the system a lot more flexible for level
designers and remove some of the confusion about how stamps worked
differently than stencils.

It will now be possible in Tile Factory 2 for level designers to do
annoying things like require pixel perfect precision or have a stencil
be a single pixel big. But I expect that most sane levels will
deliberately avoid such annoyances.

As far as scoring goes, I mentioned in the off-topic thread that I
intend to make an overall score which will be a 'breakeven time' which
should encapsulate all of the current metrics fairly well.

-D

[geironul]

I am really excited about TF2. I can hardly wait. In that vein, do you
have a guess as to when it will be complete? A day =], a week, a
month, a year?

[barryd15]

On Sep 24, 11:09 am, Jonathon Duerig <tyrec...@gmail.com> wrote:
> the game. In addition, people will be able to create custom stencils
> for a level. This should make the system a lot more flexible for level

That's really exciting news about custom stencils. Of course I
understand nothing's definite at this stage. I was wondering what
your thoughts are on custom stencils causing puzzle-code-bloat. And
what resolution?
I can see that actually building detailed multi-color artwork will be
possible but difficult with stencils, since you could only apply a
single color at a time.

Some additional related ideas:
* glass paint
Behaves like regular paint except for the way it is rendered. Allows
underlying conveyors etc. to show through. Could be called "acid" if
that seems more realistic, though I imagine it could be painted over
to restore the original shape.
I think someone else suggested this first. The obvious application is
to build interesting products which are (at least in appearance) not
composed of tiles.

* stencil inverter
Maybe spray this on a tile to invert the stencil? Not very realistic
I know. Or could stencils be inverted before spraying? This might be
useful for example to take a stencil designed to paint a crown, and
instead use "acid" with an inverted stencil to dissolve everything
except the crown shape.

* item capture
The level designer makes a small factory and pauses the simulation,
then clicks a product their factory has produced and is given a code
for it which allows them to place it in a custom level, either as a
goal or as a regular item behind a copier (see side note below).
Basically the gist of this is that the designer has some means to
extract items from their factories and use them in custom levels
(maybe insert it into their item palette). This allows any item that
can be produced in the game to be placed in a level, without having to
invent a special editor. Then the level author can pre-make items if
they don't wish the production of the item to distract the player from
what is already a complex level.

side note:
* goals as regular factory parts
Couldn't goals be implemented as a special comparator/consumer part
with the reference item/mosaic behind it, inside a red zone? If the
easiest way for the author to specify their goal is to build a
factory, they'll do a solution before the problem is completed. Maybe
this isn't too useful though.

> It will now be possible in Tile Factory 2 for level designers to do
> annoying things like require pixel perfect precision or have a stencil
> be a single pixel big. But I expect that most sane levels will
> deliberately avoid such annoyances.

I think you should adopt this as a principle in your design of the
game: Give the level designers plenty enough rope to hang themselves,
because you never know what awesome things some might do with it, and
we can all learn who makes good or bad levels (especially if there's a
catalog with a rating system).

For the same reason, I really think it would be nice to provide many
alternative parts which sometimes accomplish similar tasks in more or
less-convenient ways, establish a default/recommended set to be
available in most levels, and let the level designers take
responsibility for game balancing and difficulty if they alter this.
Having more convenient parts does destroy some of the trade-offs and
difficulties which make an engineering game interesting like real
engineering, but maybe some people don't want their level to be
focused on that kind of challenge.

> As far as scoring goes, I mentioned in the off-topic thread that I
> intend to make an overall score which will be a 'breakeven time' which
> should encapsulate all of the current metrics fairly well.

Nice ^_^. I hope there aren't any unforeseen problems.

[SaintPeter]

I've been thinking about this a bit and, despite my prior statements
against new parts, I may have one that would enhance the game.

Part Name: Edge Detect
Behavior: Emits a single pulse upon detection of a rising or falling
edge on the next clock tick, configurable at part placement.
On placement the part would default to a "rising edge detect", but
could be "Rotated" to be a falling edge detect in the same way the
rotator pieces are.
Symbol:
Rising Edge - [ _/¯ ]
Falling Edge - [ ¯\_ ]

Why: Right now it is possible to make your own rising/falling edge,
but it takes 4 parts to do. I am thinking that there a lot of puzzles
that are normally only solvable by a memory chain, but which might
become more solvable if you could use an edge detect. It becomes
easier to trigger a series of events when you don't have to worry
about repetition.

I think it also makes it easier to make a counter circuit.

One possibility might be to make this an option for the MEM part, but
it probably makes more sense to have it as its own part.

[SaintPeter]

New Wire Mode Issue
There are two issues with the enhanced wire mode, both having to do
with the z-axis ordering.

1) The shortest wire should always be on top, with the longer wires
unerneath. This is critical because if you have a tight cluster of
parts with interconnects, any longer wire will go OVER the shorter
wires, obscuring them completely. It can be very hard to see the
edges of the arrow tips, and sometimes impossible. If the short wires
were on top, there wouldn't be an issue.

2) The highlighted part's wires should always be on top. If you have
a bunch of wires going to the same part it can sometimes make it look
like there is a highlighted wire there, or obscure a highlighted wire
significantly.

Side note - The purple wires are not quite as transparent as the other
wires, which adds to the confusion.

(Still love the game, can't wait for TF2!)

[barryd15]

I basically agree with SaintPeter about remaining issues in wire
mode. Here's my view:

* still difficult to see arrows connected to moused-over part
Ivan also noted this issue and suggested highlighting connected parts,
but I thought somehow making the relevant arrows easy to see was more
sensible, since we also have to distinguish between incoming and
outgoing arrows.

* drawing many arrows slows down the game
This is most obvious in test mode, but also a bit in wire mode.

My own suggestion is, in wire mode, to only draw arrows connected to
the moused-over part, and in test mode, experiment with drawing only
active arrows, maybe as a configuration option.

[barryd15]

Re: jumping to specific simulation step by simulating without
rendering

I'm really curious about the viability of Ivan's idea to jump straight
to a specific simulation step by quickly calculating (without
displaying) from the beginning up to that point.
Of course it all depends on whether the calculation for a reasonable
number of steps can be done in a short moment or not, and it's so hard
for us to guess about that.

I imagine you could then pause the simulation and return to edit mode,
and it would remember the step number. After making some changes, you
return to simulation mode and it's still paused at that step. If you
made some minor cosmetic changes, or changes that only affected a
small aspect of the simulation, then the resulting state should only
have changed a little. But if there was a critical change somewhere
in the past, the state might be completely different, and the user
would then either reconsider their change or jump back a bit to try to
figure out how the differences arose.

To avoid problems calculating too many steps, you could allow the user
to interrupt the process.

[SaintPeter]

RE: Jumping to a specific Simulation Step
This was why I proposed a "Breakpoint" part. I believe that it, in
conjunction with "Super Speed Playback" (Ah-la the Production
Competition version of the game) would enable you to get to where you
want quickly. I don't know if the super speed playback renders each
simulation step, but I am certain that doing the calculations in
memory would be faster than displaying them all.

I just know that it's a real pain to debug the 3rd instance of an
object that takes 2+ minutes per to make. My hope would be that you
could "run to breakpoint" and then do it again to get to the next
step, and so on. I'm not sure that it makes sense, in the context of
the game, to be able to say "run to this simulation step", because
that's not really exposed to the player. I do sometimes think it
would be nice to see a running timer showing elapsed time. That would
help with optimization, since you could make a change and only run the
affected portion.

It's funny, this is basically a "programming" game, so now we're
asking for programming tools. This game is easily Turing complete.
Where is my compiler? Where is my "Hello World"?

[Lokilotus]

for me, the random part would be very useful. take my randomiser. i
NEED a random part to finish it. otherwise it's just an algorithm.

[Iván Nieto]

What's the state of the bugs in 2.0? I don't know if this has been
reported, but here it goes:

eNoFwVEKACAIRMFXuQai+BXd/6Q1M2AHnoxKUzuzfKkvxJECMPp9EhwB5Q==

Note how two white paints are destroyed for each purple.
But if the white path is moved one space to the top:

eNoFwVEKACAIRMFnbQai+BXd/6Q1Y7ADT6xSameUT/WFOEsBiH4fEekB4g==

Now two purple paints are destroyed for each white, so it's not an
element order issue anymore.

[Jonathon Duerig]

I've finished the last update for Shattered Colony that I plan to do.
I won't be working on that any more unless some hideous bug crops up.
Next up is getting Tile Factory 2 for everyone (any suggestions for a
good subtitle, btw?). I did want to point you to the map sharing for
Shattered Colony. I am currently planning on doing something pretty
similar for tile factory. Though I may add some kind of screen-shot
feature as well if I can think of a good way to do it.

http://www.kongregate.com/games/duerig/shattered-colony-the-survivors

If you want me to do map sharing differently for TF2, now is the time
to say something.

Also, there was some confusion about what "Configuration library?"
meant. This would be a 'library' where you could save particular
configurations of parts and drag them onto the map. The idea is
similar to what is in Konstructor. The idea is that you would be able
to re-use patterns more easily.

I'm also looking for more ways to make 'memory chains' less
attractive. Including the wire count might help.

I will let you guys know how far along I get in making Tile Factory 2
each Monday until it is finished.

-D

[Neebat]

"Tile Factory 2 - Floored"

In regards to memory chains, some random elements would help, but a
random part won't help (because you can't force people to use it.)

How about this?

Let's define some "bad" tiles. The "badness" could vary:
Simple copier failure. The copier spits out nothing and you need
to detect that and not waste paints trying to paint it.
Evil, subtle invisible defects that you can't see until you've
done some processing, like "Glue won't stick to this tile."
And some variations in between, like "This tile dissolves in
turpentine.", "This tile flies to pieces if you rotate it."

Now, let the level designer decide how evil to be. Add up the number
of tiles expected to be used in the level. For example, my Color Test
has two 9-tile goals, for a total of 90 tiles. So your copiers are
going to need to produce at least 90 tiles. The level designer would
pick say, 5 copier failures, 1 glue-immune, 5 non-rotatable and 2
turpentine. Then, those 13 bad tiles would be mixed in with the first
90 tiles produced by the copiers.

The key thing is, if a tile fails, it needs to fail spectacularly, in
a way that destroys it, or we won't have the tools to detect it.

[barryd15]

On Oct 5, 1:27 am, Jonathon Duerig <tyrec...@gmail.com> wrote:
> good subtitle, btw?). I did want to point you to the map sharing for
> Shattered Colony. I am currently planning on doing something pretty

...

> If you want me to do map sharing differently for TF2, now is the time
> to say something.

From the Shattered Colony Kongregate page:

"Try a random map after the campaign.
Browse player made maps."

Are those supposed to be links? I couldn't figure out how to access
existing player-made maps. I saw the "Download Map" option with the
"Enter URL" prompt, but didn't have anything to put in there.
I was hoping that a URL could be given to the web browser (eg. by
simply clicking a link) and that would load the game and the map, no
user effort required. Seeing an "Enter URL" prompt, I'm hoping that's
an *alternate* way to supply the URL to the game, not the only way.
If I could find the player-made maps, hopefully it would be more clear
to me how this system currently works.

> Also, there was some confusion about what "Configuration library?"

...

> similar to what is in Konstructor. The idea is that you would be able
> to re-use patterns more easily.

I assume Konstructor = KOHCTPYKTOP. Very nice.

> I'm also looking for more ways to make 'memory chains' less
> attractive. Including the wire count might help.

I appreciate mem chains more when I think of them as the analogue to
Codex's manipulator programming. It's much nicer how they are
implemented as an in-game part rather than coupled tightly with the
game engine. This opens the possibility of having alternate control
methods with equal status, by simply introducing new parts which the
players/designers may choose to use or not. If you have a relatively
flexible framework in which you can program new parts with minimum
effort, adding parts to support new control schemes (and other
alternate ways of doing things) like this will allow the game to be
expanded over time. Just make sure the puzzle/solution formats allow
parts to be added to the game without affecting the existing designs.

For making mem chains less attractive, I feel that you will need to go
no further than including wire count.

[Jonathon Duerig]

On Tue, Oct 5, 2010 at 3:04 AM, barryd15 <barry....@gmail.com> wrote:

> From the Shattered Colony Kongregate page:
>
> "Try a random map after the campaign.
> Browse player made maps."
>
> Are those supposed to be links?  I couldn't figure out how to access
> existing player-made maps.  I saw the "Download Map" option with the
> "Enter URL" prompt, but didn't have anything to put in there.
> I was hoping that a URL could be given to the web browser (eg. by
> simply clicking a link) and that would load the game and the map, no
> user effort required.  Seeing an "Enter URL" prompt, I'm hoping that's
> an *alternate* way to supply the URL to the game, not the only way.
> If I could find the player-made maps, hopefully it would be more clear
> to me how this system currently works.

I'm talking about inside the game. Start the game, click "New Game",
and you are presented with the play games menu.

Just above 'Download Map', there is a 'Browse Maps' option. Click
'Browse Maps' inside the game. That will give you a list of the 5 maps
rated most fun. Then you can click next/previous or change the sort
order.

Sorry for the confusion.

-D