Euphemismus
patreon
Weekly Standup 52
🕑 Added 2022-02-12 19:05:35 +0000 UTCCover image: Look at those beauties, waiting for some final checks before being uploaded.
Important news - So yeah, I can give you a definite release date now: tomorrow!
---
TL;DR:
Things I got done this week: Finished programming and writing work on v.0.12, worked a new image of Ari to existing content, lots of testing and bugfixes. The automated tests have become much more sophisticated by now.
Things I want to get done next week: Release stuff tomorrow, and after that I'll have my usual post-release-listen-for-bug-reports-and-fix-stuff-and-take-it-easy-week.
---
Monday - Programming: Planned functionality for new h-scene is done. Writing: Final draft on new h-scene is done.
Tuesday - Programming & Illustrating: I changed the new ari-sitting-down-with-teacup image from a CG (just a flat .png) to a new layeredimage (several .png files layered on top of each other). This allows me to use the tea-sitting image anywhere in the game, instead of having a specific version for a new scene in v.0.12. Most of my time on this was actually spent finding every instance where Ari sits down for tea and implementing appropriate sitting image variants for all of them. Testing: Started creating automated smoketest for the heart path to test all of the places which have teatime scenes
Wednesday - Friday - Testing and bugfixes, testing and bugfixes as far as the eye can see, including several fairly important fixes for v.0.11 which the automated testing brought to my attention. I built up the automated tests a lot over this time - they can use the save/load functionality to test different permutations of scenes or dialogue trees, instead of just linearly playing through the game and calling that a day. It doesn't catch everything, but I can at least make sure that no crashes happen.
Saturday - Final comb-over for v.0.11 and v.0.12 and building/uploading/doing all that so that everything would be nice a ready to release tomorrow.
—
The automated testing stuff isn't paying off in dev time yet, but it is definitely paying off in avoiding sanity damage. An enormous part of testing PT is regression testing (ie. "making sure that all of the stuff that worked before still works") and the amount of it needed can get pretty ridiculous - if I find a minor issue and make a single tiny fix that's probably ok, I still require proper systematic regression testing to be sure that I haven't introduced a crash onto the critical path or something.
Being able to automatically do that regression testing is fantastic for keeping my brain focused on important stuff and I've been finding a lot of fairly severe bugs in v.0.11 (which I fixed) thanks to this. I'm hoping v.0.12 is better, as well, since it has had the advantage of automated testing since the beginning.
---
So, I'll be posting updates tomorrow, stay safe until then! :)
