The presence of bugs doesn't necessarily indicate useless software. If tests are failing (or flaky...), that's probably something to look at at some point, but that doesn't necessarily mean it's the highest priority to look at. In most places where CI gets deployed -- at least in commercial environments -- there seems to be a goal of making test failures a non-maskable interrupt.
I did admit this was controversial! But it fits in with a more general view that there are a lot of tools which make good servants but poor masters. Auto-builders are a good thing, partly because (as you say) they can help to clarify what is required to make a build, and partly because (especially for dependency-heavy software, which seems to be the norm nowadays) they can help catch things quickly when the dependencies shift beneath you. Making them a hard gate on releases seems a little too close to making the tooling your master, though.
(Somewhat separately, I also worry about CI acts as a hiding place for complexity. Sometimes it reaches that point where nobody knows how to make a build without the CI tool any more. Then local testing and debugging becomes difficult.)
One of the most annoying things to deal with in a CI workflow is flaky tests
I guess another controversial opinion: this is a problem with the idea of CI (or at least how we work with it and what we expect to get out of it), rather than the idea of randomized tests.
A valid answer might be "nowhere". Code review is a hot topic at the moment, but I'm pretty sure a lot of software I've found useful over the years was created with little or no formal review, and I think we should at least consider the possibility of leaving trustworthy developers to work without constant oversight.
A side effect of this is that if your individual productivity goes up, your real wage might go up, but if your whole industry’s productivity goes up, it’ll be mitigated by this effect.
Articles telling us to focus on team, rather than individual, productivity should probably be viewed through this lens.
There's an argument for valuing ten years of Silverlight experience -- especially if it includes some gnarly projects where the framework isn't really holding your hand that much -- over ten years of flitting between web frameworks but never really getting much beyond implementing the stuff that shows up in the tutorial.
I think "going into the weeds" (and not being afraid to do so again, in another context) has some value independently of any specific patterns or tricks that apply to a particular technology.
I'd certainly agree that if a quick fix starts to accrete complexity, that's good evidence that it's time for a rethink. "ad-hoc modifications", I'm less certain about. How often is this happening? How much effort do the modifications take? Would the modifications still be needed given a "done properly" version (my experience: often yes). Would they be substantially easier?
It can go either way, and I'm certain not trying to argue that quick fixes should never be replaced -- just that trade-offs should be considered.
Over the years I’ve learned few things are more permanent than a temporary solution.
This is something your hear in a lot of places. Often (not always) it seems to be spoken with regret. I think we should be quicker to celebrate than a problem has been solved (and seems to be staying solved!) without worrying so much about whether it was done "properly".
Indeed, I think the frequent success of temporary solutions should be seen as a challenge to the prevailing orthodoxy that software developers should be striving towards more discipline and process. Instead, we should be looking seriously at why quick fixes and cowboy coding work so well, and asking whether we can apply these more widely.
This can be taken to extremes. Some years ago, a popular historical novelist told me something along the lines of "I often talk to people who say 'I'm writing a historical novel', at which point I'm wondering whether this could be a potential competitor, '...but I'm still doing the research', at which point I realize that there's not a lot to worry about." Looking into background and what others have done is valuable, but it's pretty easy to let it prevent you from ever taking action yourself.
It's a good set of questions to keep in mind, but I'm not sure it always leads to a feasible course of action.
What I like is working independently on a gnarly problem for some extended period of time. That explicitly includes spending the odd day "spinning my wheels" without being immediately being pushed to ask others for help -- because bumping my head against stuff while working alone is the way I like to learn stuff.
This way of working seems to be under fairly vigorous attack in favour of "everything needs a team", and I'm not quite sure how to work around that. Just pointing out that I'm getting things done does not seem to be sufficient.
However... skimming through some of those, they actually look like pretty typical "work on a team" positions (e.g. "...technical guidance and mentoring of Computer Technicians with focus on accurate completion of work orders and execution of standard operating procedures..."), not "this cupboard has spares, this cupboard has rations, see you in 6 months, good luck!".
Any hints for finding something that's closer to the second option?
- weekly 10% self-directed and dedicated research time
- additional research time with sign-off from a lead
These are both really great ideas, and have the advantage that they work for people who prefer to work through new stuff in private. Sometimes, organizations seem determined to combine learning/exploratory stuff with team-building exercises (A lot of Hackathon-type end up in this bucket, for instance), and to my mind that can end up severely limiting the amount of quiet experimentation that happens.
A coworker from a previous job said something along the lines of “How much work is getting done by teams debating tabs vs spaces?”
That's a good question, but having agreed that such debates don't contribute much, you've got two ways to go: either enforce a standard, or "live and let live" (perhaps supported by strong-ish module ownership). I think it's unfortunate that the "live and let live" option seems to have largely fallen by the wayside over the past few years.
I'm actually pretty certain this approach, with the TODO file in VC, would work fine at least up until 4-6 developers, as long as everyone hits a basic level of competence -- but in practice, someone always suggests a more structured tool well before that point.
A big one is that managers are de facto judged by the size of their team (and, for more senior ones, the size of the tree underneath them, too). It seems to be really, really hard to have middle management in an organisation without "maximise headcount" becoming an unofficial goal.
Edit: Employee happiness is an interesting one here. It's quite believable to me that being seen to cut down a few tall poppies might have a positive impact on (mean) employee happiness in some cases.
I actually quite enjoy the true mysteries that nobody has touched for years. It's a situation where you're expected to slink off into a corner and swear quietly for a few days, and nearly-certain hero status if you pull it off.
The more frustrating situation it needing to jump in and make a one-off fix to something that is at least notionally "modern". These can be just as impenetrable, but you have to hold your tongue much more (especially if some of the developers are still around), and there's generally a lot less kudos for sorting things out. Personal "favourite": debugging stuff that's virtually impossible to build locally "...because we've got this great CD pipeline".
I did admit this was controversial! But it fits in with a more general view that there are a lot of tools which make good servants but poor masters. Auto-builders are a good thing, partly because (as you say) they can help to clarify what is required to make a build, and partly because (especially for dependency-heavy software, which seems to be the norm nowadays) they can help catch things quickly when the dependencies shift beneath you. Making them a hard gate on releases seems a little too close to making the tooling your master, though.
(Somewhat separately, I also worry about CI acts as a hiding place for complexity. Sometimes it reaches that point where nobody knows how to make a build without the CI tool any more. Then local testing and debugging becomes difficult.)