Keep a social journal of your work progress as you make and learn things.

April 2017


Finally got the #JSON encoding and decoding working for the #MidnightMurderParty editor. I’m finding that, while I really enjoy #PureScript as a language, the packages are lacking. I think #Elm has it beat so far in the useful docs and full-featured packages department, even if Elm is a much weaker language.

Tonight I finished adding audio to the MMP promo animation—at least as much as I’m going to be able to do before I have music to work with. Simulating movement through a room by automating the volume and panning of two clock-ticking clips with different amounts of reverb/EQ was fun. One for close to the clock, one for in another room, and a lot of tweaking.

I’m also a little ahead of my Camp #NaNoWriMo goal, which isn’t too impressive since I only committed to 10,000 words this month, but I have been writing nightly. My next goal is to shift my writing to the mornings again. I really like writing in the morning, but I don’t like waking up very much at all.


I’ve been primarily working on 3 things this week:

  • PureScript Editor. The old #MidnightMurderParty editor (my custom CMS) was written in #Elm 0.16 and is basically unmaintainable since 0.16 is so different from the newer versions. I’ve been looking for an excuse to learn #PureScript for a while, so I’ve been rewriting the entire editor in that. I’m still just trying to get JSON decoding working, though.
  • MMP Promo Animation. The promo animation for MMP is done, and now I need to soundscape this whole thing. I started editing with OpenShot, but that can be summed up in one word: Crash. I’m now using Blender as my video editor. So far, I really like it. While I pair up SFX with the video, @larouxn is working on the music. It’s coming along pretty nicely.
  • Camp NaNo. I set my Camp #NaNoWriMo goal for the month to 10k words. I’m trying to finish one #shortstory, write another, and maybe get some MMP writing in there as well!

March 2017


I started scheduling my todos in a calendar (inspired by @Kraahkan) to stop myself from putting them off for “productive” things I didn’t mean to be doing. I added a recurring “Writing” entry every morning before work to encourage both daily writing and getting up earlier. This means that staying up all night and waking up late don’t cut into work time; they cut into my writing time. So far, it’s kept me from staying up too late for no reason, and all my todos have been getting done.

Since my last log, I’ve:


Submitted that #shortstory I edited down by 600 words last week to a contest. Now I have two pending stories which I won’t hear back about for a while. I’d like to write and submit at least one more before then. Maybe I’ll work on that tomorrow since I’ll be snowed in by a blizzard and possibly will lose power again.

On another note, I’ve sunk way too much time into CodeFights over the past week thanks to @wearing2raincoats, who got me into it. Problem solving with #Haskell is addicting. If anyone wants to codefight me, add me on there!


I’ve done almost no dev since my last log, but I have finished two more books for my #GoodReadsChallenge (will probably finish a third tomorrow).

Tonight I also edited down a #shortstory by 600+ words (bringing it down to 12,200 words) so I can submit it to a contest next week. It’s kind of amazing how I never noticed all the useless words on previous read-throughs, but I guess that’s why you’re supposed to put a couple months between writing and editing.

February 2017


2 for 2 on #MidnightMurderParty development tonight! Both issues I tried to fix last night are now resolved. Turns out there was an issue with deploys, and the code wasn’t actually going live (I only noticed this tonight because an obvious behavior change worked locally but not on the live server). One of the fixes from last night didn’t need any adjustment; the other I completely rethought and came up with a better solution overall.


0 for 2 on #MidnightMurderParty development last night. I’m down to browser-specific fixes at this point, and solutions are (a) dependent on other people deploying and testing things for me and (b) mostly guesswork. This slows down my ability to dev tremendously. Out of the two fixes I coded up last night, neither worked, which is kind of a bummer. Using some alerts, we managed to figure out the error message thrown by one of the issues (getting error messages from mobile browsers is a pain). My guess about the error was right, but somehow my solution didn’t fix it. Still not sure why.

On a brighter note, I once again submitted a #shortstory I’ve been trying to get published to a contest and expect to hear back about another one within the next couple days. I never really have high hopes for publication, but I definitely won’t get published if I don’t keep trying, right?


No progress on the game engine tonight—not even a name.

I went out to a coffee shop to read and see if they sold their own beans (I just got a Chemex and a coffee grinder), and while I was reading, a short but massive thunderstorm blew through. The whole shop was flashing with lightning, rain flying sideways with the wind, and the bed of a pickup truck flew out of the truck and hit the coffee shop window. The storm only really lasted for maybe ten minutes, but I later drove back to my apartment in the rain to find that all the street lights and stoplights were out. It was so dark and rainy on my way back, I couldn’t even tell if I was on my side of the road.

Suffice to say, it’s very hard to program with no electricity, so I ended up working on my #shortstory with my manual typewriter, a flashlight, and a candle until the lights finally came back on. What a night.


As of yesterday, the #Elm game engine started to look like an actual engine instead of a mess of #ECS code mixed with half a game implementation. I plan to use this for later optimizations. It originally started as an extension of #MidnightMurderParty, but I think at this point it’s just become its own project. I should think of a name for it.

Changes I made on the engine tonight:

  • When initializing a game, the user can now issue a Cmd before providing the engine with the game data. This allows for basing the game data on HTTP requests and such, which I intend to use for MMP to retrieve JSON from a database as content is released.
  • With a bit of extra type magic (there’s already plenty going on), I got the engine tracking which system (as in Entity-Component-System) is currently running, along with some extra data about the system. I also added a function that will allow running one system from within another and restore the original context when the nested system finishes.

I’ve spent the last few days working on an engine for text-based games in #Elm, the goal being to use this engine to code the #MidnightMurderParty interactive segments. Turns out, modeling a game engine in a purely functional, strictly typed language is tricky. Since game objects have different properties and behaviors, modeling them as the same type (so they can be stored in a List or Dict) isn’t really straightforward.

After some research, I decided to go with an Entity-Component-System (ECS) approach, which has been a challenge as the first time I’d heard about such a pattern was on Friday. Since then, I’ve written some terrible code, refactored many times, and only now have something that looks promising enough to move forward with. There’s still a lot to figure out (cross-entity/component messaging, for example), but at least now I have an idea of how this might work.


I might have called the #MidnightMurderParty countdown page complete a little early as I spent tonight editing it and adding stuff to it. Changed styles, added social media buttons, added Google Analytics, etc. But now I think it’s safe to actually say it’s done.

Closed a bug report from the beta as well since we can’t seem to reproduce it anymore and can’t test my fix. We’re merging the fix anyway (after a cross-browser test to make sure it doesn’t break anything else) as an extra layer of security. Will have to keep an eye out for that bug, though.


I was looking for a todo list app recently with very specific functionality: todos needed to have a due date and a start date. After all, if a task is going to take me a week, I want to see it in my Today list a week before it’s due—on the day I should start it—not on the day it’s due. In my search, I ran into a very helpful blog post that recommended, among other apps, 2Do, which I’ve been enjoying for a couple days now. I was pleasantly surprised to find out that the blog post was actually written by @belle back in 2014! Thanks for the (indirect) recommendation, @belle! 😁

15 February

4 Heart 2 comments


I added a short blurb as well as some testimonials by the characters in the story to the #MidnightMurderParty countdown page. I’m hoping that between that and the animation which will be included on the page, it will give potential readers a pretty good idea of what MMP actually is and attract the right following. I’m still waiting for some feedback on the design, but for now, I’m calling it and saying the countdown page is complete.


Did some more work on the #MidnightMurderParty release countdown page. Added a #MailChimp form for people who want to be notified upon release and fixed a lot of styles on the page to be more responsive.

I still want to include something that makes it really clear what MMP actually is. Last time I released this project, a lot of people thought it was a family-friendly murder mystery and shared it as such. It’s not, and that did nothing good for reader retention. I’m thinking of coming up with a concise blurb like on the back of a book; the trick will be making it lighthearted and fun while also being clear about the content. But that’s an issue for the morning.


Over the weekend I did a lot of work on #MidnightMurderParty.

  • Saturday. Got together with my cousin who recorded the voice line for the promo animation with me. Did a lot of audio editing to combine the best takes (unfortunately, the best take of one part had some room echo in it—my fault for changing the mic position without thinking—so that made the editing take forever, and I’m not even sure it’s done). Sent that off to the animator. Also had a meeting with @larouxn to go over bugs brought up in the beta and lay out a plan from here to 1.0.
  • Sunday. I was supposed to go to my cousin’s Sweet Sixteen today, but I unexpectedly got snowed in—should have checked the weather yesterday, I guess. I spent the last 8 hours fixing beta issues, reviewing the animation, and developing a “Coming Soon” page complete with countdown. All in all, MMP is coming along nicely.

I’m putting aside my goal of finishing the #hakyll_portfolio_blog by Saturday as a couple things just came up that I didn’t expect:

1) The #MidnightMurderParty animation timeframe suddenly got accelerated, so the animation needs to be completed within the next 4 days. This means, among other things, that I have to dedicate more of my time to being available for quick feedback and have to orchestrate getting the voice line recorded.

2) The For Honor beta just came out, and, frankly, I just really want to play it while it’s still free.

That said, I still plan to work on the blog, just not as much as I initially intended.

Tonight, I also got feedback on the current animation, scheduled a time to record that voice line so I can get it to the animator by the end of Saturday, and wrote a #shortstory (flash fiction).


Didn’t think I’d get any programming done tonight considering I spent until 11:30pm cleaning my apartment, but I wanted to try One Quick Thing™. Two hours later, I have a very solid commit on my #hakyll_portfolio_blog and more pending sleep deprivation.

I did two things in this commit:

  • Introduced implicit parameters for my config value and two other values that are passed into a lot of different functions. Cleaned up the code quite nicely.
  • Created a Context for lists of tags. #Hakyll has a built-in tag module which is convenient for getting up and running quickly and not so convenient for customizing at all. It generates the HTML to display tags for you instead of giving you a Context with values you can utilize in your templates. Seems odd that this one bit of Hakyll is hard-coded, but hey, with enough digging through types, I got the tag data into a Context.

All in all, not bad for an accidental coding session.


Styled the home page of my #hakyll_portfolio_blog, and I think it looks pretty good. Despite this, the styles are taking longer than I’d hoped, so I probably won’t be able to launch tomorrow (today, at this point) as planned. I still need to build a deploy script too, which would be easy if I weren’t developing on Windows which doesn’t have rsync. I’m setting my new goal for the end of Saturday. I think I can finish it sooner still, but hopefully that will give me some time to polish it up a bit.


Tonight I filled out the author-data.yaml config file, which is basically a #yaml representation of my resume. The goal is to eventually generate both a PDF and HTML page by just changing this one file, thus always keeping my site and my PDF resume in sync and up-to-date. The PDF generation looks like it will be tricky (and involve me learning #LaTeX), but I rigged up the new config data with my #hakyll Context and can now display any of that data on a web page. That finishes the last of the necessary features for a minimal release of my #hakyll_portfolio_blog. Not sure if I’ll be able to throw together a design and style the whole site before the MMP beta ends on Tuesday, but that remains the goal.


Somehow I managed to waste most of the day and still get a lot done.

  • Spoke with my cousin about voicing a character for the #MidnightMurderParty promo animation. We’re meeting sometime this week to record the line, which is really exciting.
  • Made a bunch of commits to the #hakyll_portfolio_blog:
    • Changed the way URL fixing is handled to not rely on post-processing the page.
    • Added an RSS feed to the site.
    • Restructured the Config record to better handle reading in multiple config files.
    • Refactored the Contexts module to more easily handle all the various parameters needed to generate a Context. For example, to generate the Context for a post, I used to have to write something like postCtx tags categories <> configCtx config <> defaultCtx, but after doing some mconcats and introducing a new record, I can do something like postCtx contextParams and it does all of the above. Pretty neat!

Spent the whole night playing with different comment system widgets because I’m getting tired of Disqus (guests can never figure out how to leave a comment, and I don’t blame them; Disqus makes it so hard to find the submit button for a guest). I couldn’t find anything to suit the needs of #MidnightMurderParty any better, though, since I need the AJAX thread loading Disqus has. While Widget Pack has that too, its admin interface seems weaker than Disqus’s, and it bundles a star-rating system in with the comment stream, which I don’t want. Couldn’t find a way to disable it, but maybe I’m missing something.

On the bright side, my #hakyll_portfolio_blog does not need AJAX thread loading, so tonight I installed IntenseDebate on there. For the functionality it does have, it’s behaving faster and more reliably than Disqus, which is nice.


Tonight I added tags and categories to my #hakyll_portfolio_blog with some custom behavior outside of what #Hakyll already offers. That was the easy part.

I spent the last four hours trying to get the Contexts for posts and projects to contain one another such that a post page can display any projects mentioned in the post and a project page can show a list of posts that discuss that project. The actual Contexts weren’t too hard to code (despite having to delve into the docs and fit types together), but the recursive dependencies were a nightmare. I spent at least two hours fixing just that.

In the end, I did get it all working, but I’ll save just how I did that for a full-length blog post. For now, I think the hardest parts of the wiring are done. There are more Contexts and config files to rig up, as well as an RSS feed, but I don’t think anything I have left to do will be as complicated and confusing as tonight’s features. All in all, it’s coming along.


Did some more work on my #hakyll_portfolio_blog tonight. I made some progress on the design/css, made some custom routes (they convert default routes like /content/posts/a-blog-post.html to /posts/a-blog-post/index.html to get rid of parent directories and hide ugly .html extensions), added some functionality to direct all links to the “clean” routes instead of the index.html ones, and got started on a #yaml config file which I’m successfully reading in and parsing (turns out this is pretty easy in #Haskell with the right libraries). All in all, a pretty good start after the initial setup yesterday.


I’ll make this a quick post since it’s late. Since the #MidnightMurderParty beta is now in progress, I’ve shifted gears a bit and started working on my #hakyll_portfolio_blog again. Really, I restarted it from scratch since the first version was a mess. Maybe I’ll be able to launch a minimal version before the MMP beta ends–that’s my hope, anyway. Other than that, I started reading my third book of the year and got some #writing done on a #shortstory.

January 2017


Fixed both of the remaining #MidnightMurderParty front-end bugs with one tiny commit!

My suspicions were correct: when I had to mess around with the renderer after Chromium broke itself, I accidentally regressed some code. The bugs didn’t become apparent until I zoomed the page to 175% in Chrome (or did anything in the mess that is Firefox), so it took a while to notice, but the fix was simple once I guessed the problem.

The column-width property doesn’t support percentage values, so to make it dynamic, I manually alter a stylesheet with #regex. Prior to the Chromium fix, the width and height were set this way as well, but then I changed them to 100% since that seemed less sketchy than live-editing more of a stylesheet. Apparently, that was even sketchier and caused the bug. Changing it back to the way it was fixed the offset pages bug (which I was trying to fix) and somehow magically fixed a different bug in Firefox (not complaining).

I also added some stuff to the testing doc.


Yesterday was a mess. What was supposed to be two quick commits for #MidnightMurderParty and the rest of the day off became a 5-6 hour dev session where I discovered some pretty nasty bugs, at least 3 of them fairly serious front-end issues. @larouxn had to create a new endpoint on the back-end for one of the features I was working on, and while I was testing that, I found out that our release scheduling wasn’t working properly either (it was releasing content that wasn’t yet scheduled to be released; I’ll leave that log to Nick).

Long story short, I did close the last two #GitHub issues, but now I have to spend tonight fixing bugs in the most volatile part of the application and hopefully not introducing new ones in the process. So much for not fiddling with it until after the beta test!


Tonight, I continued working on the documentation for the #MidnightMurderParty beta test. I filled in a lot of the functionality details from an earlier version of the testing documentation, but a lot has changed since that version. I’ll probably finish the draft of the new doc tomorrow.

Also, between @wearing2raincoats and me, we’ve picked up 3 or 4 more beta testers. Still looking for a couple more before the beta rolls around next week.


I’ve gotten back to #MidnightMurderParty over the past two days.

  • Yesterday I did most of the research needed to write Episode 4.
  • Tonight I started working on the documentation for beta testers. I wrote up a broad checklist of the functionality testers will need to verify and reviewed it with @larouxn. Tomorrow, I plan to add a larger section detailing each item on the list.
  • Also tonight, I:
    • Set up the domain for the beta test.
    • Compiled a list of pre-beta todos (all small things that I can do this week)
    • Integrated some stuff into the MMP Slack for fun.

#MidnightMurderParty development is finished!

Okay, maybe not entirely, but @larouxn and I just closed and merged in the last remaining development issues tonight. There’s still the beta test, which I’m sure will reveal some bugs, but until then, I have no reason to open up Sublime Text or fiddle with code. It feels like finishing that first draft of a story: yes, there is editing ahead, but right now I’m just really excited to have completed a first draft!

There’s still non-development stuff to be done for MMP (#writing, editing, configuring services like #Mailchimp, etc.), so I’ll still be logging about that, but for now, I think I’ll take the rest of the night off… maybe read a book, or just try to sleep off this cold.


Between being up early for a work class and the sudden cold weather over here, I’ve managed to get sick. Trying to rest up, but I’m bad at resting, so instead I worked some more on #MidnightMurderParty. Finally seem to have gotten the Safari bug fix right, and I’ve cleaned up a lot of the URL hash navigation code. After regressing some of it, I even added some comments so I wouldn’t repeat that mistake.

Earlier in the week, I joined a 24-hour #shortstory contest, so I wrote the first draft of that tonight as well. I might do a little more editing tomorrow before I submit, but for now I need to think of a title and get some sleep.


Added a pretty big feature (code-wise) to #MidnightMurderParty tonight: URL navigation. Now, when the user gets to a new segment, the URL hash will change to reflect the selected segment. Readers can share these URLs to send someone to that part of the book (much like the share links inside the book). When the the user changes the URL hash (think forward and back navigation), it will navigate the user to the relevant segment as if they selected that segment in the table of contents. There was briefly a recursive navigation issue, but I think I fixed it. Still needs more testing and cleanup, though.

Still got that Safari flexbox bug, it seems. The CSS autoprefixer did something pretty janky and converted flex: 1 1 0 to -webkit-box-flex: 1. The fix for Safari was supposed to be using -webkit-flex: 1 1 0 instead of -webkit-flex: 1, so tossing in -webkit-box-flex: 1 did nothing to solve the problem. I might just have to add -webkit-flex manually.


Tonight I:

  • (hopefully) fixed a Safari flexbox bug in the #MidnightMurderParty Reader. Since I don’t have any Apple devices handy, I’m waiting on @larouxn to confirm.
  • Moved over to Google Domains
  • Edited the #shortstory I wrote the other day
  • Printed out a QR code which I now need to scan to turn of my alarm. I put it in the kitchen next to the coffee maker.
  • Started reading another book (2/25 for my 2017 #GoodreadsChallenge)

I also wanted to set up my new phone tonight, but I’m waiting on a case for it before I switch over. Amazon Prime 2-day shipping was due to show up today, but neither item I ordered came. One said in the shipping tracker that it would arrive today before 8pm all the way until nearly midnight. Now both items have an estimated arrival time of 2-3 days from now. Apparently 2-day shipping became 4-5 day shipping somehow.


A lot’s happened over the past couple days. My graphics card died, rendering my computer unusable, so I had to learn how to install a new one on the fly, which was simple but nerve-racking since I’d never installed hardware before.

I wrote a #shortstory this morning—really a flash fiction piece—starting at 12am when I should have been asleep. I think it came out well.

@larouxn and I closed a couple more #MidnightMurderParty issues tonight. We finished up the load time optimization, so now the site shows low resolution images first while the high resolution images load in the background and are swapped in when ready.

We also began planning out the beta test, which is now slated for the end of January through the beginning of February. We’re looking for more beta testers, so if you’d like to help beta test the Midnight Murder Party Reader, let us know!


I’ve got bad news and good news tonight. I’ll start with the bad. My graphics card died, and I can’t even get into Windows now. Guess I’m going to Best Buy tomorrow to pick up a new one.

The good news is that I’m #writing my current #shortstory on a manual typewriter which, as it happens, doesn’t require a graphics card to work. I got a good amount written tonight.


Closed three #MidnightMurderParty issues tonight, one of them being the app-breaking bug introduced in the last few Chromium versions!

I actually hadn’t planned to delve into that nasty Chromium bug until next week, but I was too curious to hold off. I ended up debugging on and off for a few days until I narrowed the bug down to this: if you use CSS columns and an overflow value other than visible (i.e. overflow: hidden|auto|etc) on the same element, all columns after the first few will render but become invisible as if visibility: hidden is applied. The point at which columns become invisible seems arbitrary based on seemingly random things like the height of the container or even the size of the viewport (which should have no effect on the fixed size container).

Tracking that down was a nightmare.

In the end, I fixed it by moving the overflow: hidden to a parent element of the div with columns and then changing all the dependent logic.


Let’s see if I can get back to daily (or at least near daily) logging again. Today, I:

  • Closed two #GitHub issues for #MidnightMurderParty. Both were small aesthetic things related to the user resizing the book, but because of some race conditions that could occur between #Elm and #JavaScript (and did occur last time I tried this), I’d been putting the issues off for a while.
  • Got some #writing done on my current #shortstory—about 700 words on the manual typewriter, which I’m enjoying quite a bit so far.

Now, I’ll probably read for a little bit and maybe play some Osu! or Overwatch. Something that starts with an ‘O’ anyway.


Hi. It’s been a while.

While I was gone, Chromium went and entirely broke—and I mean broke, as in unusable broke—the #MidnightMurderParty reader. Yeah, one version of Chromium, it works. The next, it doesn’t. Since this breaks every Chromium-based browser (Google Chrome, Opera, Yandex, you name it), Midnight Murder Party is suddenly unreadable in all of them. The elements render, but after page two, they are all invisible as if visibility: hidden is set (which it isn’t). The only thing I’ve managed to figure out is that it’s somehow caused by CSS columns (each page is a column). @larouxn and I were planning to have a beta test next week, but that’s shot until we figure out how to get around all these broken browsers. In the meantime, I’ve added Google Analytics and better error handling/reporting.

Unrelated to MMP, I spent the holidays with my family, set a 2017 #GoodreadsChallenge of 25 books, picked up my manual typewriter from the shop, and have begun #writing a story on it!