Triaging my someday/maybe list

After yesterday's post about my someday/maybe list, I decided to be a little more proactive about things. I dumped all of my current items into an org-mode file and assigned custom properties for estimated size, area of responsibility, and a future column for the action I want to take.

Using org-mode's column view, it looks a bit like this:

Column view of triage list

The columns are configured with the following settings, and I added properties for each column to make data entry quicker:

#+COLUMNS: %60ITEM(Details) %10AREA(Area) %8SIZE(Size) %8ACTION(Action)
#+PROPERTY: AREA_ALL Personal Work FreeSoftware
#+PROPERTY: ACTION_ALL Do Delegate Defer Delete
#+PROPERTY: SIZE_ALL hours days weeks months years

I wanted a rough idea of how large tasks are, so I didn't try to be too accurate with the size column; using hours, days, weeks, months, and years gives me enough detail to decide on actions to take. I also assigned an area of responsibility (e.g. work, free software, etc) so they can be tied to larger goals.

Finally, I went through each item on the list and asked myself a couple of questions:

  • Would I be okay if this thing never happened?
  • If it does happen, do I want to be directly involved, or am I okay with just putting the idea out there?
  • I if I am involved, would I be okay hiring someone to do some or most of the work?
  • Do I want to do it all by myself?

These questions are similar to GTD's 4 D's: do it, delegate it, defer it, or delete it. If I wasn't sure about any of them, I left the "action" column blank (aka deferring it).

A lot of the list is still deferred, but it was a helpful first step and it eased some of the discomfort I'm feeling. Going forward, I want to make triaging my someday/maybe list part of a monthly review. Weekly feels a little too often, although that might change as the list gets smaller.


Someday? Maybe not

The someday/maybe list is an integral part of Getting Things Done. It's where things go when you might want to do them at some point, but now is not the right time.

It's a really useful tool, but even with regular reviews it can get pretty darn large. My list is around 250 items, but that doesn't include things like books I want to read, games I want to play, random ideas I've had, etc. It's probably closer to 500-600.

All of these items range in size; there are things I could do in an hour, tasks that would take months - or even years - to complete, and everything in-between. It's a little overwhelming to look at it all in one go.

One thing that stood out is that I have a pretty decent handle on the someday part of my list, but I'm struggling with the maybe side.

There are a lot of things on the list that I'd really like to do. There are also things that I'd like to do, but that are outside of my current abilities. And there are also things that sound like a good idea but are probably not so great.

Even if each item only took a day, it would still take me nearly two years to get through everything. If each item took a month it's closer to 50 years. Ouch.

Right now I'm trying to accept that maybe not is a part of the someday/maybe process, and that most of these ideas will never come to fruition.


Regolith Linux

The first GNU/Linux distro I ever used was MEPIS. It was a Debian-based distribution that came on a live CD and used the KDE3 window manager. As a first step into GNU/Linux it was a bit of a revelation for me, although I didn't switch completely for quite a few years.

Since then I've tried out a number of different distributions and window managers. I've been using Xubuntu as my primary operating system for 10+ years, and although it's my firm favourite I still like to try out new setups every now and then.

I'm currently trying Regolith Linux, which is a little different as it uses a tiling window manager instead of draggable windows.

Writing this post in Regolith Linux

Most of my work takes place inside Emacs - where I'm comfortable splitting windows - so a tiling window manager seemed like a logical step.

What I like

Feels snappier than XFCE
Everything from keyboard response times to applications starting feels faster to me.
Keyboard controls are fast
Switching to another workspace can be done with a single key press of super + <number>. There is also super + tab for switching windows one-by-one, and super + ctrl + space which opens a prompt for searching by open window name.
Low distraction
Part of my motivation for switching was to reduce the amount of distractions on my screen. So far it's working very well. This may be because I have to remember what's active on each workspace, so I open less applications.
Easy to test
Once I installed Regolith (via Synaptic) it was available to choose as a session type during log in. It didn't break anything that was installed or replace any of my default applications.

What I don't like

Steep initial learning curve
I'm still learning how to configure everything the way I want. It's all done via config files, which is something I like, but finding what to modify takes some digging.
Unlearning habits is difficult
This isn't Regolith's fault, but there's a lot of muscle memory for me to unlearn. Most things are the same for me, such as launching apps via a launcher instead, but having windows stored in workspaces isn't my usual workflow.
No desktop icons
I know, I'm that person that stores stuff on their desktop. I normally keep shortcuts for current projects on my desktop for quick access, but Regolith doesn't have a desktop (as such) so I'm having to work around it.

Overall I'm pleased with how it's going. I really like how easy it is to split windows side-by-side, which is something I do quite often when working on client websites. I don't know if it'll ever take the place of XFCE, but for times when I need to be more productive it looks like a winner.


My nineties development setup

I got a little nostalgic after looking at PADD a few days ago; there's a large part of me that misses how I used to write software. I'm happy with my current setup, but I feel I was a lot more focused when using a single-tasking operating system.

After digging through some old disk rips I found two of my oldest projects.

The first screenshot is the source code from Sonic's Adventure, which I think I wrote around 1996. It was developed in Power Basic on the Atari ST and was a whopping 455 lines long.

At this point I was learning that line numbers were optional. Apparently I liked weird variable names too.

Text adventure development in Power Basic

Next up is some of the source from Shining Online, written in STOS on the Atari ST. The STOS editor is not the friendliest I've ever used; it doesn't use GEM so there's no scrollbar or way to view the code without list ing the whole thing. I'm shocked I was able to get anything playable built.

STOS development

And finally, I spent some time setting up a more modern environment to develop ST software in. It uses Teradesk as the desktop shell and Pure C for editing & compiling C software. I have a feeling it would fit nicely on a Raspberry Pi 400 with a bit of tweaking.

New Atari ST development

It was fun getting back into the technology that I grew up with, and I'd like to try my hand at writing some ST software now that I (mostly) know what I'm doing. It might end up as one of my 2021 goals.


If I


"If I" is a one-man comedy special by Demetri Martin. The central theme of the show is organized around the following quote:

The un-examined life is not worth living.

It's very introspective and there are more than a few places that I can relate to. Particularly the fourth meaning of "if" when Demetri talks about using a points system - something I wrote about in "Keeping a Progress Log".