3 ms·
Loved the post and the art. This is genuinely going to be helpful for me to share this with my colleagues and management. > From an engineer point of view, th
by kiawe_fire 4y ago
Loved the post and the art.
This is genuinely going to be helpful for me to share this with my colleagues and management.
> From an engineer point of view, this may sound trivial, but the truth is that your career depends heavily on how your manager, and their peers, perceive your impact and your work.
> I have to admit my track record of convincing management about the importance of my work based on this particular idiom has been… spotty at best. I still believe in the concept, though.
The above are from “diagonal contributions” but to me also apply to rake collecting.
But I have to wonder, why is this the case in software engineering?
We go through K-12 school being told to make the world a better place and have a positive impact. Most religious teachings impart the idea of helping others.
In nearly every profession or job role from waiting tables to building houses, a common complaint is when someone doesn’t “help out” when needed.
So why in software engineering is the expectation warped such that “collecting rakes” to make sure others have a better chance at successful contributions often seen as rabbit-holing, or procrastinating, or just flat out wasteful?
And especially so when done diagonally across teams? Why am I expected to only own my own pieces and not improve others?
And most importantly, why does it feel like my reputation and career still hinges on the quality of my work and how it interacts with others’ work, if I then feel I need to ask for forgiveness every time I collect rakes or work diagonally?
Something feels culturally wrong about this.
- vlovich123 4y agoThere’s a lot of rakes. Some don’t bother anyone. Some are well-managed by others and don’t need your attention. Others you may not have the skill to pick up. Yet still exist others that shouldn’t be picked up now. Knowing which ones you should pick up, how many, when to timebox it etc is a skill in and of itself. I’ve never felt unrewarded for that but: a) I already work very quickly. I try to pick up rakes in my spare time (ie my main work is blocked for some reason) b) I generally try to make sure no one else is tasked with picking up the rake c) I identify the rakes that are relevant to my work (I’m generally not fixing rakes in the Linux Kernel) There’s another problem and this one has to do with organizational pain. Most organizations are set up to not listen to your external opinion. So if you pick up rakes there the other manager will be relieved of solving the problem while your manager isn’t getting enough of your focus on your current job which isn’t helping anyone (the other manager needs to hire someone and won’t even if you’re telling them they should) Both sides of the coin have merit.