There will be more updates coming soon. My current rate of work is about 1/10th to 1/20th of what it was while I was working on VQ full time, and I took about 1 year off from VQ for other side projects, but I am currently back at it with all my spare time.
No worries, I understand where you are coming from (really). More than once I have found myself in the position where my day job took too much out of me to do anything else (even before I had a kid). It really comes down to finding a job where working 40 hours per week is acceptable, and you will not be guilt-tripped into working more hours, and ideally reducing your commute if you have one.
Times around rounded to the nearest half hour, and give or take for a given day.
Mon-Fri:
- 7:00am or 7:30 am: wake up .
- 8:30am i've prepared breakfast, packed daughter for school, packed myself for work and working out, daughter is brought to school (school is conveniently 3 blocks away) .
- 9:00am I have commuted to work .
- 5:00pm leave work .
- 5:30pm-6:30pm workout .
- 7:00pm I have arrived back home .
- 7:30-9:30 dinner (if I am hungry), chores, get daughter to bed .
- 9:30 - freedom!
- 11:00 or 12:00 - go to bed .
Weekends are not generally that free for me, but sometimes I trade off with my wife to free up some time. Usually I am busy doing work around the house (i.e. I recently built some fences, installed a sprinkler system, leveled out and graveled a side yard to put in a shed, etc).
Without caffeine I feel exhausted. I get almost exactly 300mg every day, and also try to time eating such that I don't get tired (food coma).
Its a misleading statement, so calling BS is fair :) I am not juggling babies while programming.
By "fulltime" I mean I spend about 40 hours/week in my day job, and about 40 hours per week taking care of my daughter. Breakdown: 1 hour per morning getting her to school, 2.5 hours after work. That is 17.5 hours during the weekdays. I spend probably at least 20-24 hours on the weekend with her on average. (Note that there are 112 waking hours in the week)
However (!) note that I can multitask many other things while taking care of her, particularly chores and yard work (she often plays outside with me while I am doing construction work or yard work). She is also often self-sufficient, and often entertains herself by drawing (we try to minimize the amount of TV she gets). Also (as mentioned in another comment) my wife simultaneously watches her during some of this time so its a bit easier (although as long as she is awake it is near impossible for me to work unless I am out of the house).
About doing a great job: you are absolutely correct. I do not excel at my day job (in terms of putting in overtime), but I am performant enough during my shift to keep them happy. And I do not excel at being a dad, but I consider myself a better dad than many I know, FWIW :)
Yep I suppose by "fulltime" is open to interpretation here - in this case I mean during the hours my daughter is not at school, I share half the responsibility of my wife in taking care of her. In spite of this, I still feel like I could do more and feel a bit guilty about that.
I've found caffeine to be sufficient, too scared to try anything "harder" - but I suspect many hustlers do not have wives and children, or if they do they are crappy fathers/husbands - I personally know some people like this.
I often do 6 hours, but not by intention, just because I get in the zone, but then my next day is totalled (have to go to bed early). I shoot for 8 or even 9 hours when I can get it. I also drink coffee in the evening sometimes but try not to exceed 300mg of caffeine per day, but I am weird in that I can sleep even while caffeine is running through my system. Mostly, I try to compact time as much as possible to free up a contiguous chunk to work on my own stuff, which is usually about 1 productive hour per day if I am lucky.
I do not have time for friends, as one might expect (unless you count my wife and/or her friends as friends, or "twitter friends"). Which is ok by me, since I am pretty antisocial in the real world. :)
- I work fulltime
- I commute 30 minutes each day during the week
- I am a fulltime dad
- I exercise 5 times per week
- I walk my dog 2x daily
- ~10 hours/week on chores/yardwork/construction
- I still manage to have a side project.
Its definitely doable, I think you just need to be a little bit mentally ill. :)
Note on the usage of lookup tables:
Color transitions are hard - their example here looks pretty good but in general tweening between two color spaces in a linear manner can produce unwanted "washing out" or dull in-between states. The problem is not just within the transition of individual colors, but how the relative values of nearby color affect our perception. For example, see this illusion in action:
https://twitter.com/AkiyoshiKitaoka/status/10284735661933158...
A strong concept and implementation will overcome art limitations. Just look at the success of Rimworld for example, which has sold over 1m copies - or of course Minecraft (150m copies sold).
I only used an Amiga twice growing up (did own a C64 though), so I can't speak to the experience much, but I do believe they were amazing machines from what I have heard.
QfG I (the original) was the first game I ever played on an x86 (and the first game I played on a color monitor). I cannot understate how amazing that experience was, and feel slightly sad that newer generations will not be able to experience the same leaps in technology and game design. :)
In terms of physical beauty, my favorite algorithm is the voronoi diagram. The 2D variant is rather boring, but the 3D version looks great and is capable of representing many natural phenomena from cellular division to material fragmentation (a common use is procedural generation of rocks).
median would definitely be better. As noted, median for 2017 was 2k units sold per game, but I lost the rest of that data. Regardless, average still carries some meaning. :)
great point! would be interesting to see the average sales drop-off over time. Typically games have sales spikes during the early release and then later during sales, but it tends to drop off over time. "Classic" or "must play" games, however, like Portal, tend to sell pretty well over their entire life.
Good question, I dont have the data on me anymore (I extrapolated it from Steamspy, and I don't subscribe to it anymore). To get a (potentially) more accurate picture, I would instead look at all games that sold more than X units. I think if a game sells about 20k units, its probably a more serious contender (not that some good games have not had extremely poor sales).
Graph in the blog is number of games released. My data is actual unit sales (approximated), which is important data as well. In other words, if more games were released on steam, but they were still as profitable per game, it would not be a huge issue. But the trend clearly shows the average sales per game going down. However, there is also a huge uptick in shovelware, so this could be a problem as well.
I wrote this controversial tweet on the matter a while back (embedded here for convenience [1], controversial because it uses averages!)
Which might make things seem more or less bleak depending on your perspective, but an important note from this data is that the 2017 median sales were about 2000 units per game (yikes).
Still working on it actively, although in the past two years I also dabbled with two other side projects (a node js / web gl / electron side-scroller game and a compiler). Progress is very slow due to real-life constraints but at least steady now. I will be presenting the newest progress soon. :)
From the paper: "We use the rasterizer as a potentially-visible set optimization to iterate only over pixels
for which rays might intersect a voxel, and then execute a small ray tracer in the
pixel shader. That is, we “splat” billboards that give coarse visibility and compute
exact visibility in a pixel shader. This works for any pinhole perspective projection,
including eye rays and shadow rays, so we use it for the shadow map rendering pass
as well"
Neat - I used a similar technique [1] in the last iteration of Voxel Quest, but I only rasterized single points/pixels then raytraced them into filled cubes (was not efficient other than on fill rate). However I did not invent the technique (I'm not sure who did, but I first got the idea from talking to Florian Boesch [2])
I recently wanted to pour some slabs for a sidewalk on part of my house. A 16 sqft slab (4' x 4'), to my surprise, ended up taking 750 pounds of concrete/cement and hours of effort to mix by hand. Building concrete walls is much harder and requires a complex form unless you are just laying bricks. Choose your battles wisely!
I have to optimize my time to waste as little of it as possible and try to create the largest coherent chunks so that I can ramp up on whatever I am working on and get something done. This often means taking care of the stuff I have to without any sort of procrastination (work, chores, etc).
I have a kid, a dog (that needs to be walked 2x per day), a fulltime job, I workout at least 3-5 days per week, and spend most of my weekends doing construction, but I still make slivers of time for my side project. (FWIW) :)