Comment by andrewingram | Hacker News Reader

Comment by andrewingram | Hacker News Reader

Hmm. Maybe full-stack engineering doesn't need you to believe in it?

Though to be fair it probably does, because having your entire professional community believe you don't exist creates ridiculous amounts of frustration, anxiety and imposter syndrome.

I do (with skill levels ranging from competent to highly regarded):

* UI design

* Data architecture

* API design

* HTML/CSS/JS/React/GraphQL/etc

* Front-end accessibility (not the best, but I don't just do div/span soup and call it a day)

* Front-end performance (i'll spend weeks optimising the hell out of everything I see in browser devtools if given the chance)

* API performance (see previous).

I tend to draw the line at devops and sysadmin stuff, mainly because I don't find it interesting. But for everything listed, I do it because I enjoy it. And there plenty of aspects of engineering professionalism i'm not great at, which offsets the fact that I apparently have an impossible unicorn set of skills.

Honestly, I wish I didn't and was happy doing just a subset. But if I'm spending my time doing just design, I get frustrated that i'm not coding, and if I'm just coding i'll be dreaming of design. Also, the context-switching is hell. I wish my interests would let me specialise, but they don't.

FWIW, your listed skill sets look very front-end heavy to me.

So basically you're a front-end guy who can do some basic back-end stuff. If that's full-stack then 80% of developers out there are full-stack.

I apologise, it was 1am and I miscommunicated, I listed a subset with a very front-end heavy bias. I transitioned from back-end roles to front-end in the second half of my career. I also do (have done):

* Recommendation engines (less relevant now a lot of it is easily achievable with tools like Spark)

* Data processing pipelines

* Search indexing, optimisation, and relevancy tuning for millions of products

* Actual database model definition and business logic (possibly where the confusion was, i kinda meant to say everything between database and request/response layer, when all I really said was data modelling).

* Weird stuff like integrate Django’s model layer seamlessly with RPC frameworks. Essentially you could define “foreign keys” that worked across service boundaries and they’d work transparently and decently optimised.

* Systems architecture. Defining service and communication boundaries, message queues, pub/sub etc.

There’s probably a bit more I’ve forgotten about. Does this address the imbalance a bit?

My point is that if you define full-stack as merely having the skillset to do these things at an intermediate/senior level, then there are plenty of people who fit the bill. The issue is time management, I don’t think it should be someone’s job to do all these things, quality will suffer regardless of ability.

eh, throw it up on heroku

[x] devops

I mean yeah, as long as I can get by with Now or Heroku all is good. But that approach doesn't work indefinitely.

Reply on news.ycombinator.com