Why asking super experienced ppl to bootstrap your project is the best decision you will ever make?

Posted by Cute_Activity7527@reddit | ExperiencedDevs | View on Reddit | 34 comments

Ive been woking in this industry for over 12 years. For some those are rookie numbers, but there is one rule I think has the biggest impact on your overall success as a software company.

You have to start your project with the right ppl. Smart and pragmatic ppl that understand trends in IT. Ppl who can distinguish bullshit and fad from real value.

Those ppl can quit after a year or less, but it does not matter as much.

Good foundations mean life or death of a project.

Its better to pay double for few ppl who know wtf they are doing to start new project than to hire more medicore engineers, even if supposedly you would go faster.

This mantra has proven itself for me over and over in many companies.

But for some reason unknown to me its like rocket scients to some and seems many many managers.

Thats it, nothing more, nothing less.

[-]

sheriffderek@reddit

I'll often hire someone as a mentor or contractor (domain or framework specialist) to talk over architecture decisions at the start of the project. Making sure you've got a solid plan and a few very experienced people can agree on it -- is priceless.

[-]

squeasy_2202@reddit

Another good move is if you're using a framework, find well structured public repos that use your framework. Look at it's structure, ci, conventions, dev tools, etc.

[-]

sheriffderek@reddit

Yes. That's good. But if I can go straight to the person who wrote those - and talk to them about it, I'd rather have that. It's funny though, sometimes the person you think will be the biggest expert isn't. I've worked with Library authors who actually didn't have as much opinions on architecture as you'd think. I guess my age is starting to show though / when the experts I'm talking to are 25 and I'm 43.

[-]

squeasy_2202@reddit

Fair with, fair enough. I personally just look for what's common across multiple codebases in the same ecosystem. When I see the same things recurring I know if found the sauce. I've also seen enough good projects to know when I'm seeing another good one which helps immensely.

[-]

DorphinPack@reddit

This is a fantastic idea. Any dos/donts you've picked up?

[-]

sheriffderek@reddit

It's been so different each time, so - not really any clear dos/don't

But I can tell you about a few times. One time I wanted help with Ember and Ember data and so I hired the person through the Ember meetup who wrote a book on Ember data. Straight to the source! I've hired people straight from a given framework core team. And I've reached out to some YouTubers who were deep into a subject. And more recently I've hired someone via mentorcruise to get a first Laravel setup going. I usually keep it pretty loose. Do they do mentorship? What would they be willing to accept for some time with them. If they're a good fit, then I can work with them to create a reasonable outline of what they'll do. Sometimes it's a meeting once a week for a month or two. In my case - it's always been a unique situation each time. But spending a few grand - or even 10 or 20 for the right things at the right time is going to be well worth it.

ericmutta@reddit

If you are being managed by someone who doesn't write code, many miseries will follow because they simply don't understand the craft enough to understand why certain things need to be a certain way.

DirtyOught@reddit

This summarizes my beliefs so much around FE web dev. I’m a FE engineer. Shits easy when you just set it up right.

Thank god most FE are mediocre at best and can’t bootstrap a project to save their life. I’m making a living off helping right the wrongs of this stuff.

Oh you thought those strictNullChecks were annoying and turned them off? Now you have hundreds of null pointer exceptions in prod. Yea those pesky lint warnings about critical react patterns just so happen to be “always wrong”. No you have massive react lifecycle issues. Oh you decided to “build your own custom component library”. I know you’re having fun. I once too wanted to have fun. But now you have a core component library that’s a mess while simultaneously being the UI backbone of 3 major company apps. Congrats. You’re now fucked.

FetaMight@reddit

To be fair, everyone has React lifecycle issues because React has a long track record of making lifecycle stuff clunky and impenetrable. Maybe the 6th ground-up rework will get it right

True.

But inexperienced FE deciding to useeffect and rely on side effects for everything is just making it impossible for the future.

thy_bucket_for_thee@reddit

If these are the types of issues you are dealing with "legacy react" projects you are clearly not dealing with anything hard.

Confident_Pepper1023@reddit

I agree, but I find it somewhat funny that in this post about smart investments you can't be bothered to type out the whole word "people", as if you're saving something by omitting the consonants.

nedal8@reddit

y use many letter when few letter do trick

thr0waway12324@reddit

Y uz mny lttr wen fw lttr do trick

Kaimito1@reddit

Y u m l w f l d t

xAmorphous@reddit

Those were vowels that were omitted sir

Thank you, kind sir :) somehow I mixed up the two terms.

Lol no worries

oe ee oe a ee oie i

BeerInMyButt@reddit

are vowels a smart investment? OP is bearish on them, do you have insider info?

edgmnt_net@reddit

A lot of business is just interested in a way to pump money, because money's been cheap and losing its value if it sits around doing nothing. Consequently, a lot of stuff is just unremarkable and it's just horizontal scaling of efforts, hence feature factories and sweatshops.

I do agree that there are huge efficiency wins in doing something right. I'll just say that the problem goes a bit beyond. Businesses don't really have very serious plans from a technical perspective. They probably see tech debt as another form of debt and as leverage. They want cheap and right now.

Ultimately these projects do experience high failure rates and increasing costs over time, as well as bubbles popping, so paying a higher apparent or upfront cost may make sense to build something better. But you probably need to change how you do business, because you can no longer just throw hundreds of features together, you have to build upon things. You need to have an actual idea of what you're doing.

Acceptable-Milk-314@reddit

But Joe over here says he can do it too and only charges $25

failsafe-author@reddit

My neighbor has a kid in college that will do it for the experience.

budding_gardener_1@reddit

better lay off your entire dev team and get them to do it, the shareholders are gonna love that

jonathon8903@reddit

I fully agree with this! A good foundation makes it extremely easy to jump from feature to feature without making too much spaghetti.

circalight@reddit

They have experienced enough trial and error through their careers to not be intimidated by it.

dnbard@reddit

I agree with what you say but I often see that project is boostraped by some freelancers for peanuts and then more experienced people to be hired to fix all this mess.

wwww4all@reddit

Sounds like you never did any kind of startup or bootstrap work.

Even experienced people can mess things up bad starting out.

Hindsight is always 20/20.

Do whatever vibe code to get that mvp, accrue tech debt as needed to survive and expand the runway.

Complaining about tech debt or how things should be done the right way are all luxury musings, that happen after business takes off.