Hacker News Reader

Comment by vonmoltke | Hacker News Reader

> I'd say it goes deeper than this: there's a greater emphasis today on the candidate being productive from day one.

Which drives me crazy. Nobody can be truly productive on day one. You have processes to learn, even if it is just an informal "way we do things". You have existing code to become familiar with. You have team dynamics to work into. It takes time.

Personally, I cringe whenever I read a job advert mentioning that the candidate will "push to production on day one". Neither company I have worked for would allow that, let alone expect it. Nor would any company I start in the future, should I ever manage to.

View on HN

It's not like you push anything important your first day. You spend two hours doing some simple bug fix that would usually take you 15 min and then spend the next six hours learning how the CI and deployment pipeline works, usually with someone holding your hand.

CI and deployment pipeline?!? What are you a spoiled millenials thinking... we build local, and copy the whole folder over RDP copy/paste onto the production server.

You may laugh, but I've seen that far more often than an actual CI/CD process in place...

Don't forget the only documentation is a printout someone remembered they had in a drawer.

^ Maybe we've uncovered what companies are trying to signal when they say you can push to prod your first day :)

I've been at more than one place that does this (one just had a sys admin be the one to do it, after it was user tested) :(

Both were pretty entry level positions.

I think that you don't get it about pushing to production on day one. These companies do that to because they have resilient systems, tests, and tooling, and they push to production all the time. Pushing to production on day one is a way to introduce new hires to the continuous deployment methodology. If you're afraid of pushing to production, there's a problem.

I don't think it's the fear of pushing to production on day one, but that a day one employee would not yet be able to produce anything worthwhile enough to push.

A very well tested codebase with resilient systems and tooling takes time to learn, it will take time for a person to become productive within it - and that time is a lot longer than the 4 or so hours you get on your first day after the legal paperwork is done ;)

So pushing on day one - regardless of dangers it poses to the system - seems like pushing to production for the sake of pushing to production.

On another note though - I've worked at companies that have adhered to best practices and done remarkably well with continuous deployment and testing, but the risk of a pushing code to production is never zero. The intent of testing and resiliency is to reduce the odds of catastrophe and increase your confidence in your systems - not to make you overly cavalier.

At my current job, we aim for a new employee to push to prod on day one not because we want the employee to begin churning out valuable work immediately, but because we want them to quickly become familiar and comfortable with the development, QA, & deployment workflows and tools.

The ticket that gets deployed is often as simple as a typo fix or a small CSS tweak, just to illustrate the whole process. The risk of deploying to prod isn't zero, but since we deploy many times a week already, the risk of deploying a typo fix is pretty low to us.

It also serves as a test for the existing employees: ideally those workflows are smooth enough that it all Just Works on a freshly-set-up environment and account, and we are comfortable enough with our rollback strategy that (after reviewing and testing) we are not afraid of merging in and deploying our new employees' code.

I believe they think that. I don't really believe that they do.

I agree that in the general case it's very unlikely for your average developer to be truly productive on day one, but I wouldn't say that nobody is. This is a large part of my value prop: I spend the first day or two getting infodumps to get up to speed, but even during that time I'm building up plans of attack to get done what I'm being paid to get done (and to me and to the people paying for me, that is productive).

I'm a consultant/contractor (depending on the gig), though, so the social dynamic is different and it's less about integrating into a team and more about understanding the exact problem domain.

It gets to me as well. I thought it might have just been my lack of skills, but I've had this conversation across several fields with recruiters who have said similar things.

The best description I've heard is that it's a spectrum from "100% training provided" to "productive on day one." Older recruiters tell me that decades ago the spectrum was too far in the other direction -- willing to hire anyone, and piling on the training. But now it's too far the other way, dropping good candidates who aren't able to perform independently on day one.

Reply on news.ycombinator.com