These questions are about a shuffled deck of finitely-many playing cards. Each card in the deck is either red or black. We don't necessarily assume that there are equally many reds as there are blacks. The physical order of cards inside the shuffled deck is linear.

A "block" is a largest-possible contiguous set of equally-colored cards in the deck. For example, if the red/black cards come up (red, red, black, black, black, red, black, red, red, ...), we have a block of 2 red cards, then a block of 3 black cards, then a block of 1 red card, and so on.

1. You could determine the number of blocks by looking at each card in sequence, comparing it against the previous card. Can you design an algorithm to determine number of blocks such that, at least with high likelyhood, the algorithm need not examine every card?

2. Now we're only interested in whether the number of blocks is even or odd. What's the most efficient algorithm?

Both were solved by Proxima in IRC, but I found it interesting enough to give everyone a try. :-)

-- Simon

For this kind of scripting, the host game exposes functions that can alter the game's state. Scripts by level authors may call these functions, often depending on conditions the authors make up. Lua and Python are commonly used for this. You could roll your own replay-like language if embedding full-fledged programming languages seems like overkill.

Arbitrary scriptable events can introduce solution twists you're forced to explore first. This is welcome in action games, where your knowledge of what's happening is limited anyway.

But I feel that such tailoring clashes with contemporary Lemmings level design: You should be able to solve most of a level by pausing near the start, then look and think. Hidden traps and scripting would seriously screw with that idea.

Clones had scripted events. Terrain would move if your Clones reached certain areas.

-- Simon

Play game. Press [ u ] to save replay from game. Automatic filename is Replay/<levelname>.lrp. Play same level again. Press [ u ]. Above file is overwritten. People become sad.

Expected instead: Any method to prevent clobbering the existing file.

Ideas that were brought up in the past:

• Automatic filename contains the datetime.
• Use the GUI widget by default.

If you use the datetime, avoid filename-untypical chars like ':' in the filename. D Lix uses the player's name and the datetime, precise to seconds, a la
AnotherLixInTheWall2p-Simon-2015-12-02-035345.txt.

-- Simon

So, Simon - you're saying you'd welcome an original/custom song mix for Lix?

Yes, with the following reservations.

The C++/A4 branch is feature-frozen, the D port will take 1-2 months until singleplayer is passable, probably again a year until the remaining things are far enough to replace.

Thus, don't rush things. I've got this feeling of negligience already when Raymanni delivers high-quality tilesets, and I don't have any new shippable code. (I have to reply to Raymanni again with more detailed comments on the sets. Lix and Neolemmix both need this feature direly: two animations per triggered trap -- a quiet motion or blinking light while idle, and the normal effectful one while eating a lem. The gory maw is obviously dangerous, but the monkey and the acid trap could benefit from a very simple idling animation.)

Ogg works out of the box; I will ship the sound effects as oggs. Tracker formats are unsupported right now, but smaller, maybe preferred. Don't go out of your way for them; do what you know best.

Something sounds a bit off at around the 1 minute mark

It sounds to me as if the sax is not in sync with the beat - coming marginally late.

Rubix: Please fix the little bug in the music, that would be much appreciated as icing on this great piece. With namida, Ramon and möbius all reporting the same bug with the sax 1 minute in, maybe you'll find their reports more useful than mine.

-- Simon

Why no music in current Lix: This is largely historical.

In the earlier years, I found that I was playing my own music in the background anyway. This fits into a workflow where you play for 1 minute at most, then terminate and work on the code again. At this time, there were no regular users who would eventually come back to me and strongly suggest music.

Thus, including music became very low-priority. There was the Dumb plugin for A4 to play tracker music, but A4 didn't come bundled with it. I never looked deeply into it. Also, I wouldn't want to ask the composers of existing music whether they'd be fine with releasing as public domain. This was the rationale around 2008, I have put off caring about it since. It's one of the problems from wanting to release absolutely everything as public domain. (It shouldn't be a problem, rant available on demand.)

I recognize that lack of music is immediately obvious to new players. Much more obvious than whatever details I care about, provided the game runs reasonably smoothly for them.

Including ogg is fatter than including tracked music. That isn't problematic yet, because there isn't yet a ton of music to include, and there will be several solutions anyway. (Include only some pieces, include lo-fi version, ..., make hi-fi version available as extra download.)

-- Simon

Do you want one for the Lix board too?

No. Thanks for the offer still. :-)

I track issues for current Lix on the self-made forum and issues for the D port on github. Occasionally, outsiders find the project on github, and several LF users already have accounts there.

-- Simon

A subforum for issue tracking is a good, clean idea. Then people should make lots of small threads, even for small issues. The board view is the bug list then, and need not be maintained separately.

(The standard solution in software development is a standalone bugtracker. I'm reluctant to suggest that though. For NeoLemmix, I believe almost all users are also Lemmingsforums users. A sub-board will be easier to post to for Neolemmix users than any bugtracker on an external site.)

-- Simon

Clam's, Rubix's, and the lemforum pack were most active in 2012 and 2013. Clam has released updates for his Lix, pack, but I haven't looked at those updates yet.

The quality of the Lix level packs is massively ahead of the quality of the core game. Clam has updated his pack, NaOH has released her forest tileset. The C++/A4 game got framestepping, and the D port looks as if it's got about half the features from singleplayer done. Despite being a massive technical jump forward, this hasn't yet produced shiny releases. It's reasonable if the cultural output has been way less in 2015 than usual.

-- Simon