10 ms·
Ask HN: What should a Principal Front end Engineer do?
I just started a job two weeks ago. I haven't done much. Created a UI component. I ran some performance tests and sent an explanation to the team.
I'm not being included in the sprints (we are between sprints, but nothing has been assigned to me yet), I am unclear if that's because I'm a principal or they just forgot about me. There are 17 people in the daily standups but only 5 of us talk.
I don't really want to be a feature grunt (been doing that 25 years), but I'm worried they simply forgot they hired me.
What should I do?
- beforeolives 6y agoTalk to people. You should be asking these questions to your direct management. They clearly don't have a great onboarding process (even more difficult if you're remote) so you'll need to do some extra work to stay on top of things and figure out how your role fits within the team and the company.
- randcraw 6y agoYep. Ask them what are the essential duties of your role. Then ask, 'But what would be really GREAT?" Though it's probably obvious to you, my first thought about describing to others how to deliver an ideal front end would be to "minimize impedance mismatch". Somehow you need to plant a flag in the minds of upper management that you know how to make an app "a joy to use", or totally obvious to use. But is that what your management really wants from the person responsible for the front end? Or is it something else, like inventing a signature visual interface to better brand the product? Or streamlining the app's infrastructure to simplify reengineering/refactoring/repair? If you're looking to create a manifesto that you can sell as a unifying strategy for UI, IMHO no company has valued front ends more than Apple, and Jef Raskin was one of those most dedicated to making computers 'obvious' to use. His manifesto ("The Humane Interface") has to be a good source of ideas on which to base your 'front end mantra', as would the book "Macintosh Human Interface Guidelines" by Apple.
- durnygbur 6y agoStay calm, use moderate language, collect paychecks.
- chovybizzass 6y agoI've been doing that for 2 weeks now.
- dogma1138 6y ago2 weeks is nothing, that’s not even enough time to learn the inner workings of the company and understand it’s business well. Many companies have an onboarding process that can last a couple of months until you are put into a full rotation, the more senior the role the more time it takes because senior roles require much more understanding and information before decisions can be made. The biggest thing you can do as you climb up the ladder is to rush things.
- o-__-o 6y agoAt one job I was up and running in 3 months. In my current job I’m just starting to feel comfortable in decision making after 2 years! This is the ramp up time to hiring, I think 6 months is a good average target. Over a year means the company has serious technical or management issues preventing a smooth onboard transition.
- mesozoic 6y agocontinue until retirement
- Viliam1234 6y agoSo now you only got 24 years, 11 months, and 2 weeks left, congratulation! Don't worry, it will be over sooner than you think.
- nullserver 6y agoDiscover what the companies pain points are. What the customers pain points are. What the blockers are. Is there any elephants in the room being ignored? It’s also quite possible they just wanted a feature jock that won’t mess things up. Perhaps that’s what is actually needed.
- ralusek 6y agoAre you a principal engineer or the principal engineer? If you are a principal engineer, they might've just used the title to try to recruit someone who isn't a novice. Either way, if you want to assume a leadership position, just start doing shit. Especially if they're not giving you stuff to do, just start taking the initiative. Are tests set up? Do they use Storyboard for components? Set that up, or create new ones. Do they have Cypress or equivalent e2e? Do they have Jest or equivalent snapshots?
- chovybizzass 6y agoThey hired me because they have no experts in their framework. I asked if there was anything I should focus on and if not I could do an a11y analysis. He replied with spiking on upgrading to 3.0
- AnimalMuppet 6y agoThen part of your job is to help others to become experts in that framework. What don't they know? Sure, it's many things. What's the first piece you can teach them (or at least some of them)? Can you work with them to implement something and help them learn that piece at the same time?
- enduser 6y ago> He replied with spiking on upgrading to 3.0 It looks like you just answered your own question about what you should be doing.
- cybrexalpha 6y agoFirst I would talk to your manager about their expectations. They might not have a clear idea either, but you should see what they say. As general advice — as a principal engineer, similar to a staff engineer[0], your role is really to look ahead and solve the next problem. What isn't a huge issue now, but will be in 18 months time? Where are the companies projects moving? Can you predict that they're going to need particular support in the future? [0] https://staffeng.com/ https://staffeng.com/
- chovybizzass 6y agoThanks.
- o-__-o 6y agoThis is precisely what I’ve found in my roles above senior engineering. Senior management doesn’t want you heads down in code unless you absolutely have to be. They want forward planning. You need to manage technically, the good part is that personality management is much much less of a distraction.
- cgdub 6y agoI have found myself in a similar role and lead a team that spends most of our time improving the front-end build tools and front-end testing infrastructure that all teams in the company use and splitting the company's codebase into micro-frontends where it makes sense. Depending on the company and product, it might make sense to focus on performance or building a shared component library. In our case, the average team's productivity was far below where I believe it should be because of dull tools. Figure out where your company needs improvement the most and make it better. If you have the leverage, 17 people in daily standups is a good problem to fix.
- tacostakohashi 6y agoI've often found that getting work to do after starting a new job is surprisingly difficult. My suggestion is to find out what your manager is doing... and try to do it for them. Or otherwise, do this with some other people in your team. Possible outcomes are that they'll be happy, or they might say "no, don't do that, do this instead!".
- nullserver 6y agoI was a recent hire at company. I complained to VP I didn’t have enough to do. I didn’t have a free moment for years after that.
- treis 6y agoStart doing code review on the more junior front end devs.
- ecesena 6y agoI'd be happy to talk to you about by next startup.
- jonas_kgomo 6y agowould like to hear about what you are working on
- ecesena 6y agoDidn't want to be overly promotional, it's all in my profile
- austincheney 6y agoIn most large companies the title Principle is the top rung for engineers aside from something like Senior Scientist that requires post doctoral education or something amazing that changes the direction of a company’s technology. As a principal your primary job is to be an engineering executive. That means making decisions, studying, and advising executives/planners, and directing the path of technology. As a front end principal you need to find the shortest path to accomplish the business goals with front end technology. The biggest challenge is that junior engineers, especially on the front end, need tremendous hand holding and resist outside a small comfort zone. That is a challenge because as a principal you set technology direction instead of holding people’s hands, which is the job of their managers. If as a principal you are too aggressive in your pursuit of product quality you will be sidelined and ignored. If you are too lax you won’t offer any value for your high compensation. This where soft skills become important. Set the bar high, communicate the importance of high standards in terms of money, and get buy-in from senior leadership. > What should I do? Document where the current product/process sucks and recommend a radical order of magnitude improvement. Write that clearly for stupid people but include strong evidence for your case. Get buy-in from those that are in a position to care about making/saving money. You don’t have to win the support of resistant junior developers if you make a strong business case. Then oversee the successful implementation and be planning the next big challenge.
- anm89 6y agoI once had a job that forgot about me in a large corporate environment. After about a month of showing up to stare at a wall I just stopped showing up. I used that time to job hunt on the clock and ended up collecting my salary for a few months at the other place.