A Little Bit Random

Today was a bit of a random day. Did some testing and bug fixing, worked some more on maps, had a dental appointment (that involved way too much blood for simply replacing a couple aging amalgam fillings), and had a meeting at work about comms/notification in the event of a North Korean missile attack, and priority/procedures for alerting the public. (and plenty of other random stuff beside.

Residential Area Map

This is the WIP map of the residential area where the player character lives. It's the connection from your apartment to everywhere else. From here you can walk to the downtown area (when it's ready), get in your car to go to the large main map, visit the corner store, visit your friends, and a few other things.

It's already got an image map set up and in the game, just have to put the code in so that you can visit it in the next release. ^-^

Artist Update

I've always been more about drawing things, than hiring people to draw things, so this is a bit of a learning experience. With the exception of AlexW95 and Shino, the artists I've contacted have been fairly slow to respond. I'll probably send out a round of emails in the next day or two to see where things stand with the artists that expressed interest in the project. 

Random Numbers

I've added a few tests to take a look at the performance of your browser's PRNG as it relates to Accidental Woman. I've put plenty of explanation into the game on the topic, but the gist is that different browsers handle the js call for a pseudo-random number differently, and the quality can be affected by several factors.

There are two main things the tests check for: 

1) result clumping, which is when random numbers generated close together tend to have similar results. This is more commonly found in browsers that have their PRNG strongly based on the timestamp at the time of the request. (Firefox is a good example of this methodology.) When the timestamp is nearly the same, such as two Math.random() calls within 1 millisecond, results can be very similar. (I haven't actually made a pair distribution test yet, as it's a bit more complicated, going with a simpler test for now.) This is really important for weighted random numbers, which work by combining multiple random numbers generated in succession.

2) Uneven distribution, which is where a set of results aren't equally likely. Chrome had this problem until version 42, when it switched to xorshift128+. A slightly uneven distribution doesn't really matter, but if it gets too uneven you'll start to get the impression that the PRNG hates you (or likes you).

Why does this matter?

Because some issues in the game can be caused by weighted random numbers not being properly weighted. Things that use procedural generation (like clothes) will start to give weird looking results. This will be especially true for NPCs, which will be created using very complex procedural generation. When I'm trying to balance game mechanics, skill checks, and object generation, it will be very useful to get a better idea of the player's PRNG quality. (Which can be affected by several factors other than just your choice of browser!)

Naming Raffle Update

Still planned for the 5th, I want to wait until Patreon finishes processing, and give anyone who got a "declined" or "fraud" error a chance to fix it if they want to. 

Release Update

Things are moving along quite nicely now, but there's still some catching up to do. (and a really odd bug that seems to result in an unending for loop under certain conditions :P) I will hopefully have the release out on the 6th, likely Friday night Eastern Time, but possibly late night/early morning.