Cutting off the Tail

If you start things early, you're less likely to be late because you have more time. Common knowledge.

Let's poke at that idea a little.

Why do you need more time?

Obviously, because you don't know how much you need. It's a defense against the worst case.

If you know how long something takes, the risk of starting late goes way down. In most cases, you already know roughly how long something will take before you start. But there's something which will take double, triple, even quadruple what you think, and you won't know which till you start.

You can get half the benefits of starting early if you can lock down an estimate. Cut off the long tail of possibility.

That still isn't perfect, because estimation isn't perfect. But it's a good enough bar to get through most of life. You can still give yourself as much buffer as you want if you don't trust your estimate.

Maybe you find out you need to block out more time and move around your calendar, or maybe you find out that urgent task you've been avoiding would actually take 10 minutes.

Point is, having a good estimate gives you benefit without the time cost. You don't even need to start, just look at the expectations and constraints and make sure it's within reasonable boundaries.

Flexibility

How much flexibility would you have if nothing had to be done early?

I look at so many things while walking between things.¹ Messages, documentation, specifications. You can't get much real thinking or work done during that time, but you can check on so many things.

I check on the constraints for many side projects for weeks before I actually get around to doing them, because then I know when I actually have time to start.

An example

I want to download a certain kind of data and keep it cached locally for easy access and better search. This isn't urgent in the slightest, it would be a quality of life improvement.

If you've ever worked with web scraping at all, this amount of vagueness should already be giving you shivers. This could take 30 minutes or 12 hours for just a functional prototype. Such is the nature of the web.

Good news though.

I've already checked, there's APIs for all the data I want including for listing where to even find it. It comes in a roughly standardized format, and I've done enough poking around to be certain I know how to download it. There's more complicated libraries as well, I'm still debating if I want to use them. I'll probably sit on that for a couple more weeks.

Zero rush.

I know most of this from idle searching and reading when I was between places or didn't want to do anything. It's near zero actual time investment. But when I actually get to implementing, I know it'll only take a few hours to get up and running.

No 12 hour tail.

I have real deadlines too.

Same principle, smaller time scale.

Unless you have exceedingly good estimates and guarantees that the task will be exactly as you expect, prioritize checking things over starting them.

You can start later, but only if you know you can.

- Rew

Nightly Notes

There was a long time in my life where everything was done three days in advance. I stopped because it was more stress than it was worth, or seemed to be worth. Since then I've developed a bad habit of going down to the wire on everything, and I'm still working out how to find a better inbetween.

It's been interesting watching the tail on writing go down. It rapidly degraded from "I need to move around everything in my life to make sure I don't forget about the new thing" to "ah, I have an hour, let's knock that out for the day real quick".

Also, hooray, it's the end of January!

31 posts and 18k words later. We made it.

We?

I don't really know who I refer to.

There is you, the probably existent reader.

There is you, the future version of me reading this.

And there is you, the old me who already wrote the last month of posts.

I am none of those people right now. I keep signing off twice. So, we.

Didn't end up doing anything special, that's fine by me.

Thought's on the whole blogging thing haven't changed much in the last few weeks. Numbers haven't changed either, I still can't tell if anyone's actually reading all of these. Somebody's reading some of them. Maybe it is the same somebody.

Hi again, if you're real.

Anyway.

I don't see a reason to stop, so onwards we go. Next month will be easier, I'm sure.

- Rew