2021-W40-2022-W01, 2021-10-05/01-08

the previus log was in early october. at that time, the three client projects i was working on for about 1 or 2 years were supposed to be finished by end of 2021. there is a crushing drive happening when things are going well on a project, or when things are shaping up nicely after months of messy head-down work, where whoever is asking for things get swallowed in by their own emotions (?), or any other kind of energy boost, and start to demand more. this has been happening with a certain cadence before any agreed upon, in-progress deadline for the projects i've been burning on.

the same happened for two, if not all three projects, some months ago and everything collapsed down, with deep scars.

who gives a fuck, here another two-three months to finish work! a breakkore mantra sound we heard for the past two years. whether this time it might happen for real, i can't tell. i know i will hand over work and then i'll disappear. i ended the last few weeks of 2021 thinking i want to be jobless for a while in 2022, unclear if that's really possible (maybe by saying a lot of nos — also, yes, other people would be happy to take on website project, i always share opportunities with people in the networks i hang out).

dae.wiki

i onboarded a new dev doing backend work: the onboarding took the longest time and we might be able to do some work this month. honestly, unclear.

by end of this month we should deliver and hand over the project as well (max early february). during november and december there's been an IT / infosec "pentest" (penetration test, poor nerds and their word choice), which has not been shared in advance, so i am working on fixing the bugs reported in there. plus we had the infamous log4shell exploit as we use Apache Solr. things have been patched promptly, ITC is run by ass faces so coordination takes forever — need to setup two new VPS as well. plus doing the actual work on the wiki. at this point, i don't mind sharing shit work, that is the utmost result of working with them for 2 years and a half.

TWLA

went through pages and pages of gdocs bugs, fixes, tasks, etc — mostly visual design related. a lot of changes has been decided at the very last, some undoing freshly done work that took 2-3 weeks to put together. aim was to deliver everything before winter break, but given the constant infux of design changes, no matter how much and how fast i was churning through tasks, new tasks would be added. i raised white flag! the result of people not understanding how websites and web browsers work, and not caring to listen to the one person actually doing the idiot work.

buycloud

the still quite interesting project got stuck due to complications in running a background task and trying to have it running purely inside python (due to simple web hosting environment); it turned out that running a cronjob makes everything easier.

discussed several times how the website should be working, lots of misunderstanding and a lack of shared language. somehow we managed to communicate enough things that the other person would understand and did move on. ah.

rush

i did some work on rush, a reading-list CLI program that provides a visual interface to add, remove, archive, and display a list of URLs (in markdown notation, ayee). as with any other note i keep, the file is under git, so it was a nice exercise to learn pygit2.

it's more or less done, so i think at the next iteration i will start using it for my "real" reading-list, without the annoyance of possibly messing it up.

tekken-cal

last night i also spent two straight hours to poke at a project i began six months ago. it's mostly a way to make a decent view for myself of my calendar data — i still haven't found a calendar program that is not just showing little squares with labels.

for instance, i would find it useful to be able to do a proper search throughout my events across several calendars, or query for certain data (eg showing me all meetings for project x, and make a total amount of the hours spent). it's also a way to see if i can lightly time-track projects, when i have a time budget for it (which would be a general good idea to start to get used to).

currently the CLI program:

reads from a list of URLs and retrieve .ics files (eg, by sharing each calendar i want to query, sort of subscription-like?)
makes a local copy of each .ics, and fallback to them in case of no network connection
reads each event inside each calendar — next it will display by organized by specific time structures

still all very basic, but it feels already useful to be able, at the next iteration, to show upcoming client-work events for the next 2 weeks, or query for how many hours i spent on x project on doing x thing.

rospo

rospo means toad in italian, and i use it to publish to my website, ah!

also made a helper shell function to create a new worklog entry, as well as adding a rospo.sh file that acts as a CLI program overview — eg you can pass arguments to run specific shell scripts.

af-etc@dis-grazia$ ./rospo.sh
rospo :: publish data to andrefincato.info

Usage:

  --add-worklog: make a worklog entry using worklog-new.sh under /worklog/
  --build-page: make a page using build-page.sh and place it at the root level
  --publish: both git-push and rsync html content over to the server

having spent 2021 trying to do some CLI programs, simply slapping some echo commands to print the CLI help, and use case $1 in to check wich argument has been passed and then run the appropriate script feels so easy and lightweight. it might have all sorts of problems and edge-cases one a different computer / used by someone else, but what's the point of personal computing anyway? writing highly custom random programs to suits you and your needs. fuck docker, fuck unix.

jalla jalla! to vegetable gardens and more!