The art of interrupting software engineers
content.pivotal.io
content.pivotal.io
This is monumentally bad advice.
> I try to find a time when a pair does not seem to be heavily focused. Some of the cues might be they are having a casual non-work related conversation or are just returning from a break
So surveill and/or eavesdrop and/or ambush. Nice.
> Hadrien Raffalli is a Product Lead at Pivotal.
Here's me thinking it's probably a more junior PM doing this. Nope, it's a Product Lead.
PMs/TPMs can be great, if they are good. I've worked with some that made my life easier. I've worked with others who really didn't.
Based on this article alone, Pivotal's PMs would make me quit.
Pivotal can deliver things to customers in spite of all this (given that it’s surely not because of all this), which sort of makes any advice published out of there fundamentally irrelevant for people working pretty much anywhere else.
I have a lot of experience with XP. In my experience, XP succeeds because of its extremeness, not despite it. It's not for everybody, but for organizations and teams such as Pivotal who are willing to understand and embrace it, it works very very well.
So I challenge your "given that it's surely not because of all this" comment. Do you have any evidence? Or are you just saying, "I don't like this, so it can't possibly work?"
The kind of "pairing all the time" kind of XP that is being preached is similar for programmers. It also works for repetitive work - building yet another web application - at scale, not at scale, not much of a difference. But programmers often get into the job because it is an inherently creative field. There is nothing like this kind of XP to suck it all away and leave them disillusioned about the craft until they escape and find a product role or even a consulting company with more empathy and kindness.
However, we had a few guys on our team that really didn't like it. They left and I don't think less of them in any way. It's a different way to do things.
As for the "repetitive work" aspect: the most successful XP project I did was a collaborative sketching application using pen input with automatic shape recognition: somewhere around 2000 (when did MS put out those pen based computers??? -- We did it for their unveiling).
The problem though is the Agile Industrial Complex and "thought leaders" who have pushed this as a religion on the industry. I enjoy pairing, and the excitement of two minds working at the same problem in tandem is an experience to be relished. But give it to an Agile snake-oil salesman, most of whom gravitated away from programming a long time ago in favor of making money training others in how to program, and they'll wield it as a weapon by prescribing the only true way to work.
XP (in its traditional sense) doesn't scale past 8 developers anyway. Anybody selling you XP for a team of 500 is someone to be wary of ;-)
There are pains and giving folks the opportunity for change/rotation is refreshing.
Software is a team sport, its a weak-link game where we must communicate and empathize in order to see clearly (often past our own egos) in order to pull one-another up.
I was fortunate that good pairing hygiene was a focal point. Without that it can be disastrous mentally and ruin the experience for everyone. It turned into a highly collaborative sport at that point.
I am a hugger naturally but when I run into those pairs I went through the trenches with, it is a brother/sister hug. Best time of my life.
There’s also a survivorship / selection bias inherent to companies with extreme views like this. Either you both (a) agree to cheerlead & display positive affection for the religion and (b) display effectiveness in the role, or else you will quit or get fired or self-select to never work there in the first place.
So in a large part, it’s literally unknowable to what degree success is in spite of vs because of these extreme practices. There can’t be evidence about it because by definition there is the confounding factor that the employees whose productivity you measure are only from one specific sub-population (likes extreme view & delivers good software). You never get to measure what people with diverse opinions about tbe philosophy would do in the role, and you never get to measure what less successful developers would have done in the extreme philosophy.
The same effect happens with extreme coaching philosophy in sports, where again after you apply the filter that the players are selected to both be very good and accepting of the philosophy, there’s no way to measure the counterfactual possibilities, so it reduces to religion.
It would be an inconvenient time for an interruption, and probably looks quite a lot like returning from a break.
That having said, my personal preference is also the scheduled 15 minute meeting.
Maybe that's the idea - "He always ambushes us when we come back from a break... we'd better stop taking breaks!"
Programmer 1: Whoa, that two hour Pomodoro was quite something.
Programmer 2: We got a ton done, though.
Programmer 1: Yeah! I hope they have the lasagna again at lunch.
Project Manager: Hi, are you getting your work done? If you could get back to work, that'd be great.
First, you have to understand the context. This person is talking about an Extreme Programming team. Speaking as someone who wrote a book about XP, here are some things that are true about by-the-book XP teams:
1) All production programming is done in pairs, all the time.
2) Everybody sits in a shared team room at dedicated "pairing stations."
3) The atmosphere is highly collaborative and friendly, and there's always a low buzz of conversation in the room. Interruptions are normal and expected, as the team's overall performance is valued more highly than individual performance. Due to pairing, interruptions tend not to interrupt flow.
4) Requirements are conveyed verbally, not through written documents. Stories are short "reminders to hold a conversation," not fully-fleshed out.
5) The team values feedback and wants frequent conversations with their on-site customers (including the product manager) to confirm that they're on the right track, particularly about nitpicky details that are difficult to anticipate in advance.
Given this context, the author is talking about how he, as a product manager, does his job of providing business context and customers' points of view.
This material is not meant for people who aren't on an XP team. It's good advice for product managers who are.
Already from the article title I started to have this reaction. As a software engineer, nothing kills focus and productivity faster than interruptions. From the perspective of a manager, I can understand - and really appreciated the respect and consideration that went into the article and the recommended approach.
However, every point in the description of an XP workplace sounds like torture. Paired programming, in a shared room, always a "low buzz of conversation", requirements not written but verbal only.. To be fully engaged and productive, I need the exact opposite.
Regardless of that issue, XP isn't for everyone, and there's nothing wrong with that. I like it though, that's why I'm back. :)
EDIT: Also, I don't know him, but our writer here seems to think a lot about how he interacts with his dev team, which I appreciate, but think maybe he got into his own head about it a little bit too much here. But I also don't really mind interruptions.
> [Stories often] get written up with gherkin-like Given/When/Then sets of acceptance criteria with additional details/considerations noted down
Very interesting to hear, especially in contrast with the "verbal only" approach mentioned along with XP.
For my personality type, I work best when there are requirements/acceptance criteria clearly specified in writing. Even better if the logic is written in a DSL that all stake-holders can create together.
> I also don't really mind interruptions
Yeah, I see that software engineers come in all colors, and some are able to multi-task while fielding questions and interactions all the while. And others, like me, need more isolation / insulation to maintain silence and solitude to be able to focus and be productive.
I suppose that's a challenge for methodologies like XP. In order to be effective, it needs to take into account the different working styles and personality types that exist in a team of any size.
I think the interruption thing is interesting - at jobs where I was solo, I probably minded it significantly more. With constant pairing, different interruptions can be handled in different ways. Two-second questions can be answered by the non-driver while the driver continues to focus. Five minute answers should probably stop both members, but having two brains allows you to bounce back into context through re-merging your mental model of where you left off.
Thinking about this though, I realized that in pairing you have internal interruptions as well - if one person needs to use the bathroom or take a phone call, the other can choose to keep working and maintain context. When the breaker returns they can reload their mental models much more easily. Or you can both take breaks and you've already got a built-in ping-pong opponent ready to play :)
Wow, I never thought of pair programming like that. It sounds nice actually, to let another brain keep the ball rolling while one is busy with something else.
It makes me realize that a big part of my reluctance (preconceived notions about pair programming) is due to having worked alone for most of my career. I never experienced the kind of "mind meld" you describe. I'll remember to keep an open mind about collaborative working methods!
- Some sort of sound barrier so that I couldn't parse the background conversations.
- To be able to decide "what is our current plan?" and to write that down. Of course you'd change to a new plan in light of new information.
Individuals doing solo work are often distracted by background noise, and even I do! But in pairs, it doesn't seem to have quite the same effect in my experience.
On the planning part, I know a lot of pairs that set up small daily roadmaps to stay on track. Totally normal, and also totally ephemeral as well. Usually an updated plan is drawn every day since the pair might change.
Yes. Its an unpleasantness I endure if I’m not trying to work through anything which is an intense use of my Working Memory.
The theory is that introverts don't dislike social interaction but unsafe social interaction. It's why you're generally more social with friends and family.
Left the company for a usual startup setup with your own Macbook, pull requests, and things.
I miss the almost zero friction collaboration.
As a pair. Either of us will point out each other mistakes and suggest an improvements to what we're doing in real time without the ping-pong of pull request.
If both of us can't figure things out. We can just scooch over to other pair workstation, sit in the other chair, use the mouse and keyboard to show part of code we have issue with, point to the monitor, and ask questions.
You might want to take a look at the original short book "Extreme Programming Explained" by Kent Beck and form your own opinions.
The main thing is that every single change needs to be vertical in nature. You are never making stories where you say, "I'm going to add this functionality to a library. Later we'll add the UI and finally, we'll add the data to the DB schema and then ship". But it means that you need to design your software quite differently in my experience.
Even just from the DB level, you need to have things set up so that you can make very quick changes safely. You need to be able to change the schema, do migrations and make sure that rollouts can be done safely -- and it needs to be done on the order of an hour or so. That probably seems impossible to some people, but it's one of the reasons you can't really take a legacy project/team and say, "From today we are doing XP!"
Similarly, you need to be able to make changes to your infrastructure quickly. So if you have some big set of microservices and each small change requires a signoff from some group somewhere: well, it's not really going to end well. You need to be able to make that change and get everything working end to end in hours, not days/weeks.
One of the things that a lot of people have difficulty with XP is that it's generally incompatible with nailing communication specs to a wall. One day, I may need to add something to the API. The next day I need to be able to change that API. If someone comes to me and says, "You need to know everything you want this API to do for the next 3 months so that we can have a sane versioning system", I'm totally screwed. I don't know what I'm going to need next week let along 3 months (or whatever) from now.
For a lot of people, this stuff is just way too different to deal with. XP comes with the subtitle, "Embrace change". It literally doesn't work unless you decide to go whole hog on it. I've been successful introducing it to some teams, but on others I don't even try.
To try to answer your question a bit more closely... it's probably not what you imagine. It's not like you have people telling you how everything is going to work all the time. However, you are constantly thinking about it. Sometimes you'll have developers say, "I want to work on X" and you'll respond, "Why? How are we going to give value with that work?" And then they think about it and say, "Well, what if we altered the work flow here. That would allow us to remove this stuff. And then if we added this, it would mean that we don't need that library any more", etc, etc, etc. And instead of telling management that today "We removed some technical stuff" or "We got rid of some old libraries", we say, "We improved the workflow for the customer which enabled us to simplify the code at the same time".
This is why I was asking - I still don't fully understand if this is a "XP thing" or not.
Doesn't this reflect some much deeper problem with the product planning or management? If you literally can't plan even a week in advance that suggests there simply is no plan, and even the most scatterbrained business environment doesn't change that fast.
I'd say customers being able to stick to a coherent story about what they want for more than a week would be a key need in any successful IT project.
Methods like XP embrace that by allowing requirements to change over time, to account for the customer accumulating more knowledge about all three areas.
It's on the engineering blog of a company trying to hire a ton of developers at the moment and it reads like the most toxic work environment imaginable to anyone who doesn't have that context, so perhaps the author and the team who approve blog posts should have made that MUCH clearer.
What is your book name? I would like to buy it.
As far as I know, there's only one type of printing, and that's a lay-flat softcover.
This practice is just saving time writing down behaviour that won't be implemented.
What do you do when people forget things?
There's still something written down, just not massive amounts.
Personally i try break stories down smaller to avoid the need for large tasks which helps. I also try minimise the time between story creation and story development.
Otherwise I just communicate a lot.
Things get differently from what I invisioned, sometimes they're worse sometimes they're better.
Also a problem that's solved by the combination of pairing and frequent communication. You don't need full context on anything yourself, as long as the combination of you, your pair, and anyone you can grab within arms reach can fill in the blanks.
Pivotal R&D has struggled with maintaining context. For multi-megaSLOC codebases with many moving parts, organic memory begins to fall a bit short.
Different teams deal with this differently. Some write design notes, some keep architecture design records. But IMO it's one of our weaknesses.
He goes through a lot of psychology and science of how people work better when they're not distracted. It seems that all of the "Extreme Programming" guidelines go quite the opposite of what the science, per Cal Newport, suggests.
He also points out that collaboration is perfectly acceptable during deep and focused work, but this doesn't sound like how it should be at all.
While I’ve never worked in an XP team, the description here sounds like it’s primarily focused on making sure the individual members of the team aren’t accidentally working against each other. For projects that aren’t reliant on the sort of insight that Deep Work is teaching us to generate, the team cohesion may be able to make up for the hit to individual productivity.
No, it isn't. A PM should be having these conversations - briefly! - in front of the rest of the team in the standup, and in more depth in a weekly planning meeting.
If a PM feels a need for updates more often than that, then that is a problem of theirs that they should work in in their one-on-ones with their manager, not something they should dump on their team and then blog about.
It's sad that a PM at Pivotal has gone adrift like this, but it's tragic that this shitty advice has made it to the company's blog.
I stopped right there. My personality is such that being social exhausts me. Don't get me wrong, I like people, I just can only do social for short bursts and then I have to recharge. There are people who get energy from other people, so this might work for them, but not for me.
When do they write documentation?
They solved the problem of breaking flow by banning flow in the first place?
I can't even begin to imagine the sheer amount of meetings and "chats" that would need to take place on a regular basis just to keep things moving along.
I admit to knowing very little about XP but this sounds like my own personal hell.
> I followed my mentor’s advice and started to check-in at least every half day with each pair of engineers I was working with
Uuuuhhh...
If a PM did that with me when I have a lot of work to do it'd send me off the rails.
This article seems more to be about "what I have done as a non-technical PM to reduce my own anxiety at the detriment of my team".
It's hard to say without a lot of detail, but I think there's something disfunctional about the team or about how the project is structured in this example. How big are the tickets? It's common for a "user story" to take days or weeks; it can say something simple like "user can submit a form" but have a lot of big system dependencies that were not well thought out (a form system, a state system, a server component, etc).
Instead, tasks must follow a technical logic (a form system first, state management after, etc) and engineers should be left to do their own thing, with minimal "checking". Something like BaseCamp's "Shape Up" [1] comes into mind, with a few bets every cycle.
In my decade of experience i’ve usually seen new PMs be extremely anxious and paranoid about timelines. As they get more experience and they see more projects that go off the rails they understand the tech speak a bit and have laxer expectations.
Sure we’ll be 2 weeks late, but in the grand scheme of things, it doesn’t matter. Don’t burn your engineers for a dumb deadline that makes little difference to the customer.
There’s a reason we have kanban boards and sprint boards. How are things going? Look at the board. The tickets are updated. Any blockers? Read the slack channel. Only interrupt in person when shit is really on fire. But even for that, the monitoring software will page the opslead.
The brutal truth of software development is that the quality of your output is going to vary with the quality of the engineers you work with. Good engineers don't need daily standups to build great products, and bad engineers aren't going to be helped by doing twice-daily checkins.
Checking in with a pair isn't a status update, it's really about collaboration and visibility. Estimates are often wrong, assumptions get tested, the team dynamic is "let's ensure we all understand this".
Stories in tracker tend to be <1 day, so checking in occasionally isn't exactly a wrong thing (every half day to me is excessive), so priorities can be shifted, or big stories can be broken up.
If my PM was constantly worrying about whether I was distracted, or working on something outside of the backlog, I wouldn't be too happy about that. As an engineer, part of my job is to prioritize my daily and weekly tasks, and to avoid getting distracted throughout the day. Unless their is a hard deadline coming up soon (in which case it is helpful to check in with the PM so you know what to cut), requesting status updates on stories on a daily basis seems like way too much. If you want a window into what I am doing, come to standup (which should be async most of the time anyways, IMO).
I've worked at a company founded by people from Pivotal, with a very similar culture, and I'm not very surprised by this post. There were a lot of new bootcamp grads and more inexperienced developers, so the company didn't really give them the same freedom that developers are used to. Everyone pair programmed all the time, even when doing basic UI work (super exhausting), working hours were fairly strict, and work was broken down fairly finely. For better or worse, developers were treated more like just people who can write code, for the most part, rather than people who are able to intelligently prioritise tasks and take initiative on things that need to be done.
It's like that episode of Seinfeld where Kramer pretends to be Moviefone. "To update the status of your ticket, press 1 now. .......... Why don't you just tell me the status of your ticket?"
For actually tracking the progression of the sprint we have a secret Slack channel that management can't see.
This way we don't have to explain to a spreadsheet monkey why the burndown chart looks like a skyline.
This makes me so sad.
The jagged burndown chart should be the perfect tool to explain management that you didn't finish the sprint BECAUSE MANAGEMENT CHANGED OR ADDED THINGS, and so visualize the cost of such actions.
1.) Treat them as engineers who are capable, able to communicate problems or expected delays, and professionals who want to grow. 2.) Treat them as assembly line workers who constantly want to screw around and need to be whipped back into shape.
I'd argue maybe a third is the startup mentality where you need to sell your life to the company, but maybe that's a mutually exclusive thing.
In my experience, the second comes from the same mentality as the old waterfall companies. Spec everything, trust no developer, micromanage, etc. Developers are a risk, not an asset, and never trust them at a management level. I would never work for a company like that, and avoid them like the plague.
Regardless, both of the views above seem to be a self fulfilling prophecy. Professional shops who treat developers as engineers produce some quality stuff and usually on time. Sweat shops have a lot of failed and over-budget projects and a high turnover.
Fred Brooks said "plan to throw one away, you will anyhow". He was mostly talking about the value of prototyping or what in some circles is called spiking.
> I suppose if you write a lot of the same kinds of software you can estimate up front pretty accurately though.
Not in my experience, and I'm a pro-estimate loony. A lot of products and problems do have fairly predictable elements and it's possible to give useful estimates (eg, add new form element, needs a new field in the endpoint, new column on the table, some validations, tests for the above, no changes to docs ...).
Other problems are closer to pure research and you just accept someone might come back with something some day.
What we argue about in software circles is how wide valley in the middle is.
I think an engineer needs freedom to be able to exercise judgement say "I don't understand this task well enough to estimate it yet."
Why? Speaking from personal experience, when I was a recent grad and got asked for an estimate, said "I don't know" and then got pressed for one, my thought process was "okay... I guess I'll just come up with a number and double it" because thats the advice I'd been given multiple times. This resulted in an estimate that had no relationship with reality and which I knew had no relationship with reality. This was not helpful for communication because...well because it was a lie. Thats the key: unless you give someone the freedom to say "I cannot yet give you an estimate", you can't know if the estimate you get is a lie.
Establishing this freedom is hard:
* One reason that it can be hard to give estimates of simple tasks like adding a new form element/adding a new field in an endpoint/adding a new column on a table is if you don't yet know your tools well. But it can be rather embarrassing to admit this. So the engineer can be reluctant to exercise the backbone to claim this freedom.
* There is a lot of moralizing around estimates. The article On Being a Senior Engineer[1] says "Avoiding responsibility for estimates is another way of saying, 'I’m not ready to be relied upon for building critical pieces of infrastructure.'" so admitting that you don't yet know how to estimate might be admitting that you should be fired.
* This is hopefully an uncommon case, but an engineer might have been unwittingly hired as a senior engineer right out of university because their recruiter told the business he'd known the engineer for years. This exacerbates the previous point.
* A business stakeholder really wants to have an estimate, even if the reality of the world is that engineer can't give them one.
* There aren't any resources for new engineers to learn task-estimation[2]. So if one of your engineers doesn't know how to estimate tasks, you might worry that they'll never learn how, since coaching that engineer to learn estimation is too time-intensive an option for most companies.
* When an engineer is asked "how long will this take?", it isn't obvious from that question that the estimation process requires first identifying the key pieces of the task and the riskiest pieces of this. That relies on knowledge of how to produce estimates: understanding that they key piece of estimation is not the number you get at the end but is the mile-markers you produce along the way.
All of these difficulties hurt the key thing required for a good conversation around estimates: trust. Trust that the person hearing the estimate knows that there is uncertainty. Trust that the engineer will flag up if they run into a pile of complexity.
[1] https://www.kitchensoap.com/2012/10/25/on-being-a-senior-eng...
[2] Demystifying the Black Art by Steve McConnell is about project estimation, which assumes you can already estimate individual tasks.
Absolutely. This is itself a valuable estimate. Very occasionally engineers will give a very high points value to indicate that the proposed story is simply too complex as written and needs to be decomposed further.
> Establishing this freedom is hard:
It absolutely is. Everything we do at Pivotal that works well is actually hard. Our cheat code for bootstrapping these practices is that we already bootstrapped them years ago. But maintaining a culture is harder than building one, which is a trillion times harder than improving one.
What I have witnessed in those hip, agile young companies is that you have to report exactly what yyou have done, what you are plannign to do etc. constantly. At least once a day, plus whatever other weekly meetings you are in. Sometimes your entire day is filled with meetings and you are supposed to fit your actual work somewhere in the breaks. And if you can't finish a task in a day, you trigger extra attention from managers and suddenly you find yourself sitting at your desk with another engineer and a manager looking over your shoulder, wasting hours of expensive time instead of letting you work.
Those are the two extremes I have encountered and they are only annecdotes. But they do serve to show that the development process chosen for a project is not tell you anything about the management style.
I have worked for another large telecom that treated its developers great that did kanban.
Currently I work for an agency that really does neither. :-D
But I think you're right, the methodology doesn't matter nearly as much as the managers.
The whole point (at least at Pivotal) is to spend more time coding/designing/testing and not in meetings.
It wasn't like this before fads like XP, Agile, Scrum etc. and micro-management became mainstream. Most of the current management done by PMs and their ilk can be easily categorized under "Bullshit Jobs" (https://strikemag.org/bullshit-jobs). The problem is that once somebody has codified their experience under a catchy title and started marketing it, people blindly follow it without understanding its context and applicability. Given that the barrier to entry to these sort of positions are so low (you can become a PM without any sort of domain knowledge!) people who might not cut it as Engineers seek these jobs to boost their pay, power and egos.
PS: The above might come across as a little harsh but unfortunately is the reality for most Engineers.
Damn... is it too late to go back to college and study electrical engineering?
Problem is it also takes the developers who are 8,9 and 10s and turns them into 5s too.
If you have lots of 2s and 3s this is probably fine. If you have more 7s, 8s and 9s it's a problem. For a start the better developers will end up simply leaving as working conditions are too hostile.
Most developers are < 5 / 10 skill wise, so these silly methadologies become stupidly popular.
Thanks.
Daily meetings are not good. They waste people time and constitute an interruption. If you can sync the whole team so that the meeting happens on the morning before people start working, that's better, but still not good as you are imposing a schedule on a team that may know better.
I just finished a deadline at 2 am yesterday. My team is sane and did not expect me in the morning today (joined the slack around 11 am). Would not have done that if I had a daily at 9:30 am. I would not only have missed the deadline but the daily would have even further delayed it. What's more, I would have spent some time preparing it because I would feel I need excuses ready in case someone starts a blame-game.
If developers time is precious, don't waste one hour per day on dailies/daily preparation.
The team I joined - 4 developers and 1 manager - goes for a coffee walk in the morning. The manager waits for us to go through email and other communication that came in overnight. Then we walk to a nearby coffee shop. On the walk, we discuss issues and what we intend to accomplish for the day. It takes about 20 minutes.
Occasionally, we take a walk at the end of the day as well to discuss issues. When the manager calls for an end of day walk he senses our energy is low.
As I get older, I like routines in the morning. They help me get my mind in gear. We go for the coffee walk. When we get back I might read a short technical article. Then it's time to get stuff done.
See, to me, neither of those is a great idea. In the morning I'm coming in early before most folks so I can get a few hours of real work done without distraction. In the afternoon I'm mentally fried and ready to do anything other than office work. The best time to get me? Schedule an effing meeting so I can plan my day appropriate, and can make sure I'm apprised of any information I need so I'm not walking uselessly into an ambush. Doesn't have to be a big, formal affair in a conference room.
I think the moral of this whole discussion is that managers manage people, and people are all precious little unique snowflakes with differences. You either create your organization around a certain class of personality with everyone on board and the non-conforming out in the cold, or you create an environment that recognizes people aren't robots and work around those vagaries. Of course this is more a continuum than a binary.
Lazy managers treat everyone the same. Lazy managers also avoid setting some semblance of professional standard, leaving the mob to sort it out.
Software engineers have largely been nocturnal, favoring later arrival times at work, in my experience. In my team's case, all the developers are late arrivals so our manager simply works with that tendency.
A startup I worked at in the mid-90's had people working in pairs: a server developer and a client developer. They didn't care what hours you kept as long as you didn't hold up the other person in your pair. I arrived around 9:30am daily. The client guy I was paired with came in around 8:30am. I always stayed later so I'd check in code before I left and leave a message for him about the progress. He tested stuff before I arrived and went over any issues when I got in. Then we'd work together the rest of the day. Other engineers came in at 10:30am, noon, and as late as 1pm. They worked it out with their other halves the same way I did.
I think wise managers work with the team, minding company-imposed limitations, rather than enforce personal preferences.
Late day meetings are great too if it works better for the team.
In my team's case, we also walk or ride bikes together at lunch. So we can also use that time if necessary, but usually we just keep it casual.
People who end up in manager roles tend to be people persons and some use very touchy-feely language. I'm personally put off by all the references to his own anxiety as just one example. That doesn't necessarily mean he's actually a terrible manager.
Part of his job is to make sure things get delivered in a timely fashion. If that's not happening, then feeling stressed by it isn't unreasonable, but the article sounds like he's just trying to manage his emotions rather than get deliverables from his people in a timely fashion.
I'm not posting this to dog the author. I'm posting this to suggest that his framing may reflect the fact that his job involves managing people rather than things.
If it were framed differently, the exact same practices might sound much more reasonable and palatable to people here on HN who are overall reacting fairly negatively to this.
>The idea that programmers are not good with people is, in my observation, a self-perpetuating stereotype
It was never true. It is a TV/movie stereotype just like the term "Hacker".
It's not really about innate social skills per se. A lot of social stuff is learned behavior anyway.
Here's the thing: as programmers we sometimes have a bad day. Sometimes a bad week or two. Sometimes there is unresolved ambiguity in how something needs to be done that we tend to procrastinate on it until it becomes clear. When you're working through an agile/XP backlog - a never-ending backlog mind you - you're expected to produce at a consistent cadence, not much different from laying brick after brick, but unlike construction, this one is an infinite pool of uninteresting work.
This is common in consulting work. Agile and XP were created by consultants to create predictability in delivery. It makes sense for that context - clients are often far removed from the consultants, there is no intrinsic alignment between their incentives, and being external people there is always an information gap in what is best for the company and whether the contractors are making actual progress.
So you need a way to smoothen this out. Transparency, constant communication, a structured process, consistent cadence, measurable metrics, and most importantly number of hours with your bum on the seat.
When consulting companies start out, there is a distinct messaging on their websites: "we don't want to be just consultants, we want to be an integral part of your team" or suchlike. See, we also want meaningful work, we want to truly participate in your business. But as consulting companies mature they realize this is not a feasible approach at scale. While external messaging might not change, internally they realize the business is about selling X number of hours in a year per consultant, with a percentage of their earnings as your profit margin. To make this work, you need Agile/XP. You need to bring Taylorism back into the world of programming. That is just the systemic incentive of this sort of business. Companies that fail to realize this and change accordingly (the idealists) go out of business very fast. Others thrive.
Within this system it is still possible to have meaningful work for developers (I've worked in such companies, and I've enjoyed it). But if you have a micro-managing PM like in this article, then better run for the hills or you'll find out hating IT and fully burnt out in a few years.
Do us all a favour and send this write-up to every XP/Agile/Scrum/Whatever acolyte that you come across :-)
Hello Peter whaaaat's happening?
Just use scrum as intended, maybe push for more detailed updates if devs are just saying stuff like "still working on x". Or just trust them and don't try to babysit.
I especially can't imagine working with this one. Imagine constantly having to revise estimates and getting micromanaged by someone who isn't your manager. The "role responsibilities" conversation would happen very quickly
I find it especially ridiculous that this person is interrupting people two times a day for fresh updates on top of a daily standup. And they even admit their technical knowledge is not good enough to actually be able to help much if there is an issue. If you are bothering and interrupting every single person on your team multiple times a day just in case you need to increment a number in a JIRA board now instead of tomorrow morning, you are wasting people's time. Leave it to the standup.
Your job as a PM is not to micromanage engineers. If that is anybody's job, that is the engineers' managers. Your job is to make the engineers more productive and make sure they're working on the right things - but not at an hourly granularity.
We're known for it, it's part of the interviewing process for Labs and most of R&D. Folks who don't want to pair generally don't apply. Of those who apply, some try it during the interview process and decide it's not for them.
That is a good thing.
Besides which, solo work is common too. Most of the field org works solo, Spring folks mostly work solo, I love pairing but currently I'm working solo. If a pair decides they should work solo, they do. Lots of pairs have a rotating arrangement where somebody different works solo for a day or two. I know of teams where they are solo in the morning to review PRs and pair in the afternoon to work on new contributions.
None of this is enforced. There is no manual, no rulebook for The One True Way. Nobody tells you how to work. The way works is: the people are nice and smart and want to do the right thing. The rest is largely details.
That... sounds like forced to me?
And as I noted above, pairing isn't mandated. Teams are free to adopt any process they please, but most folks at Pivotal like the defaults and we tend to gravitate to them.
Pairing was optional, but the team chose to do it, so everyone on the team had to do it.
It wasn't "forced" because I "always had the option to quit the job".
It was hell.
Especially to those on a visa, the only one forcing you to remain in the USA / Bay Area / x company is yourself. To put it bluntly, no one is entitled to a job, and this applies to every person. In this scenario, it's not like strangers are being pulled off the streets by Pivotal to pair program.
Suppose an employer asks a loved one to have sex with them or they lose their job. I mean, no one is putting a gun to your loved one's head to have sex with their employer, but something is clearly very wrong at this level right? Suppose they have an expensive debilitating medical condition that can really only be treated with good insurance (I have seen this firsthand where a medication can cost hundreds a month even with insurance (not great insurance but I was an intern, it was all I could afford)). Suppose they tried searching for a job with comparable insurance for years but always come up empty. Now when it's between whether or not you will go homeless with a debilitating medical condition or basically being raped, what would you choose? Frankly, I would rather be raped. I have been in the hospital with great insurance as well, when my condition was at its worst, and even morphine and norco did nothing to make the pain go away; I can only imagine how much worse it would have been on the streets with that kind of unending pain; I can only imagine never being able to find a job like that ever again because I will always be in unending pain and never be able to interview well. Fortunately, I got really really lucky in that I am an engineer who is an American citizen that can find jobs easily that dont require me to physically be there but, most people are not in my position. Back to the point, if someone is more willing to be raped than to not be raped out of fear of the consequences, that's pretty damned close to forced right?
But if I worked for Company X and then one day Thou Shalt Pair Or Else, then yeah, it would feel shitty. Labs developers pair with client developers on engagements and sometimes it feels like they've been dropped from a very great height.
Most of the time it works out fine: people are people and we work in the same industry and share many of the same experiences and passions. Sometimes it just doesn't work for everyone, which is also fine.
But insofar as anyone from Pivotal is explicitly or implicitly saying "or else!", I think they are making a mistake and hope they get feedback about it.
R&D leadership otherwise pretty much don't interfere. Their main business is explaining our strategy, keeping the lights on and teams connected and adjusting allocations between different product and component teams.
We totally did force pairing on client developers, though.
What I meant is since Pivotal is known for uh voluntary-enforced? subscription to XP religious practices you can’t really expect it not to bleed into all aspects of dev process.
Btw I have nothing against pairing in general, just not the way Pivotal does this.
But yes, pairing is very helpful for maintaining context when interruptions occur.
Moreover, I recently fell into the trap of working for an entire week nonstop to solve a really hard problem. Took a few hours off one day to clear my mind and when I came back, I had the problem solved inside of an hour. If I continued down the route I was going down before that time off I can guarantee you it would be more days wasted.
IMO a good product manager does not focus on trying to understand _why_ things are hard, but rather _what_ things are hard.
With this understanding they can either 1) work to change scope/requirements so that the bigggest user pain points are solved ASAP or 2) accept that more time than anticipated is required to solve user problems.
Implicit in all of this is the trust that when engineers say something is hard, it is actually hard.
When I see the author talking about paranoia, trust is the first thing that comes to mind.
To be fair, I don’t think that trust is easy. But I also don’t think that effective product management is possible without it
Edit: why is the trust so one way? I’m just expected to trust them, but I have to earn theirs like some kind of untrustworthy child?
As an anchor (approximately a tech lead) I made it my business to work closely with PMs on the product: giving feedback on how the technical direction was shaping up, asking for stories to be updated or broken apart, giving options for technical seams and so on. Frequently we embed product designers as well, so discussing design options and asking for modifications was common.
But when I was pairing with another engineer we did whatever was on the top of the backlog according to our best judgement and with attention to quality. If it needed refactoring, we refactored. If we felt that would be better to wait, it would wait.
The most important discussion, I found, was about chores -- work that engineers view as paying dividends in future velocity. Our view is that engineering is the ultimate owner of chores and can schedule them according to its considered discretion. The system works because of trust. I trust that product managers and product designers will be diligent in identifying user values and priorities. They can trust us to do what needs to be done to make the software to a sustainable high quality.
I remember once I was rotated in as anchor for a product that had begun to struggle a bit. In particular we had a particular giant tech debt: we had fallen behind on React versions. Worse, there were three parts of the codebase using React, in three different ways. With my peers I decided that we were going to fix it. "I think it will take one pair week", I remember saying.
It took five weeks. But nobody ever said: stop, you work for me, my way or the highway. We talked about the reasons. This tech debt would only grow as time passed. There was no good time to repay it, so sooner was better than later.
We were trusted to do the right thing, so we did the right thing.
I don't want to read any further. Surely it doesn't get worse than that!
/s
But more to the point: it slowed development. I could deal with a bit of visual glitchiness, but my job as an engineer is to create software to solve problems. Piling hacks on top of hacks doesn't do that and it's shitty to the next engineer who joins the team.
As I noted in another comment, I would give feedback to a PM that this kind of checkin isn't necessary and try to understand how to help foster trust.
At Pivotal I have worked with 9 or 10 product managers, none of whom have asked during the day how things were going.
If anything I would be the one interrupting them.
At the moment I have been on a solo allocation for about 6 months. It has let me go down a lot of rabbit holes, which can be educational. But I miss pairing pretty badly on days when I am working on straight regular development.
That said, it really isn't for everybody. But you may want to give it a chance.
If you put two random programmers together and ask them to "pair program", you'll very likely end up with two annoyed people at the end of the day.
I was lucky in being mentored into it by some very seasoned pros, and I learned the tricks of the trade that way.
However, it's like therapy: being forced to do it instead of doing it by yourself will result in a terrible experience.
I'd rather have breakfast with my family and bring the kids to school.
Harassing engineers
> why are they getting paid for it?
Engineers got their work done despite the micromanage (to be clear, I mean despite and not because. Just in case any PM or manager is reading this and thinks micromanaging is why tasks get done)
Of course, it relieves stress and stuff, but if you're THAT stressed because of the project and need constant reassurance from the engineers, you're clearly in the wrong job.
I had to lead a few engineering teams (wasn't a PM, though) and the only approach that ever worked for me was to check things out in much larger intervals, a week, sometimes even more. Actually treat engineers like adults and be an adult yourself. Let them come to you if they have a problem. Sample the progress in much larger intervals in order to reduce noise.
Micromanagement doesn't do any good for anyone.
Also, learn how to use JIRA or whatever tracking is being used. In my company managers doesn't bother to look at the JIRA board so you end up reporting the same thing five times to different management levels.
In short, be less controlling but try to help and treat people like professionals who know what they are doing.
As a developer, all I need from management is clear priorities. Let me know what the business needs, and keep reminding me what those are in case I get lost into code. Create chunks of time where the priorities don't change so I can focus. The bigger the chunks the better.
Then let me do my job please! I need to be able to trust that management is picking the right priorities, and management needs to trust that I am able to convert those into code. Trust is paramount.
If you can do that, I will be 10x more productive than any of these pseudo management techniques. It doesn't matter if we do scrumm, agile, daily standup, use Trello or some other tool. These are just implementation details. Let the developers pick whatever they are comfortable with.
Of course there were "emergencies" and things so it was more like a team standard than a rule but it was phenomenal. When I got to my desk in the morning I always knew I'd have a solid block of time to get through the important work before I had to go to a bunch of meetings and take on more.
edit: at the most, I think async are passable via slack or something similar. Otherwise all data on this can be easily done via a Kanban board.
- sync up / status is 15 mins. Any implementation discussion gets put in the parking lot
- once we've gone around the table, the rest of the meeting is handling the Q&A / parking lot we parked during the status discussion. Anyone not directly concerned by it is encouraged to leave, get out and be productive!
This gets you the double benefits of interrupt coalescing (a single regular interruption instead of 2 smaller ones where one is an irregular surprise), and preserving the quickness and relevance of the stand-up, as it's really just the first 10-15 minutes of the calendar slot, after which you're good to leave.
In my experience, the need is to provide a clear method of communication and ability to unblock one another. A Kanban board is perfect for this, as well as committing to a definition of done.
Anything else... is very wasteful in my experience.
Also if a teammate is actively working on something that is a downstream dependency of something I am working on, then I can passively learn about requirements and expectations for my own work, become aware of issues ahead of time, etc.
1. For transparency any member of the team can identify blockers in Kanban by putting them in the blocked column. 2. If you're blocked, and running into issues, you'd reach out to anyone that can help you. Why would you wait for a standup or "talk to the wall" for that? Just be proactive.
Dependencies should be known before committing to a scope of work (iteration) with the team. Planning stage (minimal time investment for scope of work definition) vs committed work stage (in progress iteration).
At Pivotal we see the PM as part of the balanced team. The PM, just like designers and engineers, will have the daily standup, weekly retros and planning meeting together. The goal of the standup is to share what of interest happened yesterday, if help or opinions are needed and discuss pairs for the day. Sometimes discussions start about topics mentioned in the article where it's crucial to have the PM there.
Standup should usually be around five minutes. That's the reason it's not a sit-down meeting.
My view is that a PM is available as a resource to clarify as needed. Meetings between the "doers" of a team and the PM should not be forced, especially daily. This leads to micromanagement and changing requirements too often.
But alas, everyone has a different philosophy. :)
My personal preference is for very fine-grained stories rather than giant chunks of acceptance criteria. The latter tend to turn into week-long wanderings and get edited over and over. Whenever I've been able to I've coached PMs towards smallest possible units of value.
Changing requirements should only happen outside the scope of committed work; otherwise you're building churn, poor features, and poor code.
This lever of "committed work" can be adjusted per team (1 week, 2 weeks, 4 weeks, and so on).
It seems you agree with my view of a minimum of 1 week before it can change.
I agree also with fine grained tasks/stories. As small as possible and as consistent as possible. They should all have clear DOD's though.
Definitely agree with smallest possible units of value (that also varies per a person).
Finally! A sane voice appears!
*status report
> My team meets for 30 minutes each morning
That is far too long IMO and IME
> We're a small team (3 developers)
30 mins for 3 devs...
The only case where I want such frequent updates are when (1) there's some P0 bug on production and it needs to be fixed / tested / deployed yesterday or (2) we're trying to pressure the person to want to quit their job or magically improve (more the former than the latter).
Going and pestering the team is only going to slow down development process and increase the likelihood of the team not delivering for the sprint.
Tracker is able to give short-term forecasts of what will be done by when by averaging the last 3 weeks of deliveries (ie, velocity).
Personally I think the idea of sprints is a menace. It creates an artificial deadline that leads to either waste or stress. Deadlines should be reserved for actual meaningful deadlines.
It's just that I've seen the evil twin version of sprints far more than the official version.
With that in mind, sprints can work pretty well, and work "spilling over" to another sprint is generally no problem. It starts falling apart if the artificial deadline becomes a forced deadline, or if the protection part is dropped.
Indeed. I got about that far before I quit reading and started fantasizing about the unspeaking things I would do to the idiot that wrote this if I ever met him. Things involving sharp blades and fire.
Especially this (joke?) is hilarious:
"I followed my mentor’s advice and started to check-in at least every half day with each pair of engineers I was working with."
But I suspect he is not at all trying to be funny and, I'd say, it is utterly terrible advice from some PM fitting right into what seems to me a work environment turned buzzword hell.
Second of all, I've seen two kinds of product managers. One kind looks at software development estimates as commitments to be contractually adhered to. The other treats software development estimates as assets/investments with risks to be managed.
Which kind is more effective? Well, which computation accomplishes more results, the one that doesn't halt and logs the same line over and over, or the one that does halt and additionally does the thing it set out to do?
One thing that I've learned over the years is that product management is often effectively a junior management position and actually should not be a management role at all.
That puts things in perspective in terms of their authority. I've ended up coaching PMs on my team more than once in terms of helping them understand what the product was (often non trivial for new PMs joining an existing team), what kinds of things are easy technically and what kinds of things are not. Even just understanding that there is low hanging fruit to pick (i.e. opportunities for them to look good) is non trivial, etc. All the good ones I've worked with were quick on their feet in learning who to talk to and who to take advice from. A good project manager does not need to be a domain expert to be effective. They just need to learn to ask the right questions.
Ultimately, knowledge-sharing between all the stakeholders in a project (including the customer) is key. PMs acting as bottlenecks can actually be a problem when they are inexperienced. I've more than once had situations where a PM asked me "can you estimate building X please because customer Foo needs this ASAP" where we ended up having a discussion about why building X was actually important, what customer Foo was asking for vs. what they actually needed and us both deciding that actually building something else is a much better way of helping customer Foo. Asking the why questions is always important for both engineers and PMs.
That's why it's important for them to be free to do that and why the proper job title is Product Owner, not Product Manager. In the organizational hierarchy this sits at the same level as the tech leads he's working with in the same team. As soon as you have engineers reporting to PMs, it's a problem and if they have no-one reporting to them, they are not managers.
That's the problem right there. How do you expect to manage a project you don't understand?
I don't think I will ever agree with handing the project to a non technical person instead of the team lead.
Get 1 tangible thing done a day. Ideally it has a visible effect (but don't make this religion- I had to rip out a cache infrastructure last week and replace it with a home-grown one and there was no visible effect other than things not breaking!)
Now, some days, this means you either don't work the full 8 or you go over 8 because the particular piece you're building just won't get built sooner (or you didn't break it up enough). But on average, just having this goal seems to be working.
If interruption would be costly, I take it to a quiet place for a few hours such as a library.
...or even more...
If you actually need to know whether something is falling behind schedule every few hours (which seems unusual), of course not everybody is going to provide updates that frequently of their own volition. But most good engineers, under the right circumstances, will have a sense for when something is not on track and can raise their concerns.
If you actually had a good level of trust on the team, you would work to get people to buy into it. Of course, it's probably hard to convince most engineers that you actually need an update twice a day, for good reason.
But why mask the expectations?
IMHO the best thing you can do is it to communicate priorities (which is your job as product owner / manager), and then if you do need time-based estimations, work them out with the team -- with the clear expectation that they are provisional.
Setting up expectations yourself, and possibly communicating them, but not down, is just a recipe for stress and desaster.
The focus should not be on time management, that's not why software fails, it fails due to bad design.
The "ongoing conversation" developers should have should be about modules, objects, interfaces and algorithms, not about user stories. And refactoring the system to better handle new features should be encouraged, not thought of in terms of "how does this advance a user story".
User stories should be input into the exhaustive design process, and then the design should drive the development tasks.
I'm expecting push back on this, I'd probably go as far to say that for a large team on a large project 1/3 to 1/2 of a Sprint should be spent designing, and the remainder of the time coding.
Writing tests fits in this "designing" bucket for me. TDD is a great way to flesh out how the system should be structured.
Is anyone able to share such "user stories"? I'd love to see how these business needs get transmitted in the real world.
Also - they're usually way longer than the one sentence referred to in these articles, right?
[0] https://www.producttalk.org/2012/04/user-stories-are-better-...
I don't know how many product people you have, but if everyone followed your suggestion at some companies, then the developers would spend all day chit-chatting with product managers and nothing would ever get done. Even with only 1 product manager, I can't imagine that spending ~40% of the day on status updates is worthwhile. (Literally have had meetings where the answer was "status is the same as it was an hour ago because we've been in this meeting giving repetitive status updates the whole time.")
What you need to do is get to know the developers themselves - how they work, how they communicate, what their domain knowledge is, etc. Then you'll know when you need to message them. And hopefully they'll know when to message you, but you'll also learn which ones might not under which circumstances.
Some will spend a lot of time planning, achieving nothing, but then produce something well-planned. Others take a quick stab, learn from it, iterate, and refine. And there are other styles.
Some try to understand things better before talking about them, others want to talk first. Some have enough domain knowledge that they can 'be the user', while others see the user as the one who keeps screwing things up and making things difficult.
All will occasionally get stuck in that cycle of "ok, it's _almost_ done" without realizing what they don't know until they take the next step. Interrupting to ask in that case isn't going to change anything. It's just something that is always going to happen and is (or should be) factored into the long-term planning averages.
The best PMs know their developers and know when to ping and I'm happy to talk to them when they do, even if (especially if) I'm not doing well. They don't pester me several times a day, so I don't need to spam filter them.
If you're really having anxiety attacks and paranoia several times a day when you're not monitoring your developers' every move, you should consider a therapist and possibly a more easygoing laid-back job. Software is stressful, but it shouldn't be that stressful, and inflicting stress on your coworkers isn't a great solution.
This smacks of high stress environment with a bunch of people who don't really know what they're doing, maybe all the way to the top.
Good developers don't work in these places. And if they do get caught here they don't stay long.
I'm not a PM but I sympathize with how difficult it is to discern where a project is up to when it comes to software, especially when you can't come up with short enough milestones...
They see this and it fits a pattern and so they are fairly skeptical.
Those responses usually upset me, because I feel that my experiences are being devalued. I read the comments and I compare them to my life of the past five and half years and they just aren't the same. It can be hurtful to have casual snide remarks made when you know that the people you work with are kind, thoughtful, intelligent, hard-working humans through and through.
But I also know that other people have had bad experiences and what they hear sounds like other things they've heard from shitty managers and colleagues, so they're cynical. If I talk down to them then I am devaluing their experiences. So I try not to, even though my personal nature is to be an arrogant wanker.
Context switches are annoying to the engineer and costly for development. Each context switch probably costs between at least 15 to 30 minutes, maybe more depending on what the engineer is working on. If that happens 4-6 times a day, you've potentially burned 1/4-1/3 of the day. I spent close to 3 hours in 3 unrelated meetings. So, 1/2 of my day was spent task switching and not really giving 100% attention in the meetings because of the constant task switches. Frustrating and don't feel like I made any progress today, on my first day back on an overdue week long vacation.
As a PM of about 4-years of experience, I understand the author's concerns and at the same time believe that he has solved for the wrong problem; his own anxiety rather than team cohesiveness and trust. The issues he solves for can be solved in different ways without such a negative impact on the team, actually in a very positive manner that increases trust, cohesiveness, and user context.
I'd first like to point out that I've made the mistakes of micromanagement that he mentions in the first section. I also quickly learned how detrimental that behavior is to team relationships, productivity, and outcomes. I'm certainly not perfect!
The way the author solves the problem of pace is by creating regular check-ins every half day. He then goes on to identify, how these check-ins can be used as open to monitor common slow-downs: engineering struggle and product context (he calls is product debt).
I'd first like to make the point that some other engineers have commented here, estimates are estimates, we are agile, you learn things along the way and need to adapt. I believe your job as a PM is not to hold the team to estimates, it is to come up with creative solution and serve as a master facilitator on your team.
So when the author says the following I can't help but notice the irony. I'd explore us all to practice what we preach. Instead care about iterations (sprints) not task micromanagement:
"As a product manager (PM), I often make assumptions as to when a particular user story will be delivered. This is painful because I don’t believe in time estimates and don’t think I should be involved in estimating the relative complexity of a story since I don’t own its implementation.
I know I shouldn’t but I can’t help doing it anyway…"
So how do you solve the problems that the author points to "struggle" and "context" without adding interruptions?
• For struggle, that is what standup is for. Not only standup, but if you create a culture of trust engineers will naturally reach out if they hit road blocks, make a slack channel for your team. It's even better if you can teach your team about red flags, things that the article says which are true to look out for: complexity is rapidly increasing, not feeling confident in the stack/codebase. For any of this to work you MUST create a culture of questions and answers. Yes, all want to look competent, but what's also great is being able to depend on your team.
• Adding to the previous point, this is why small tasks are soo important, [read this book on product flow](https://www.amazon.com/dp/B007TKU0O0/ref=dp-kindle-redirect?...) if complexity is small to begin with you can avoid that issue AND you get the added benefit of smaller iterations, meaning faster red flags. The author points this out too, this is critically important.
• For context, context is king. Want to avoid "Product debt", get your team/tech lead involved from Day 0 of your quarterly/yearly roadmap, they should know everything about the user that the PM does minus the RAW research effort. Run a Kickoff/Inception/IPM that truly integrates your entire team, every engineer should have product context. Devs will then see tricky edge cases, radiate them up to the PM and you can make a call together if the work is necessary.
All of this is to say that there are some good parts of this article, but I think that adding one more check-in is not necessary when there are other ways to solve the problems identified.
People laughed behind his back.
Yeeeeah... did you get the memo about those TPS reports?
>The engineers would give me highly technical explanation I couldn’t understand. This would frustrate me because I did not add much value and distracted them.
I have a rule which avoids this situation happening, ever, and its a controversial one, but bear with me:
* NEVER accept management from someone who cannot do your job.
This rule, hard and fast, has saved my ass so many times - and every single time I've broken it, things have been more painful than they should be.
Seriously, think about it. How can someone who isn't qualified to do your job, actually manage you? Only if 'management' is an administrative-only task - i.e. "report your hours worked - I don't care about anything else".
But if you put this rule into action and insist that your managers are able to do your job, you have a very strong, very stable organizational structure. This doesn't mean the manager has to do your job - just that they can, if they need to.
You can't manage things you don't understand. So never accept management from people who don't understand the job. And if that is not happening in the organization, you've got another problem: HR are just chair-stuffing. INSIST that this not be how your org is run.
When put in a leadership position, above all "resist the urge to manage".
I don't want a manager, I want someone who serves the team, insulates them from unnecessary distractions, removes obstacles, etc. Someone that supports a culture of humility, respect and trust while helping to care and feed for each individual's autonomy, mastery and purpose.
I've worked on some great teams and less than 1/2 of the time these leaders could not do my job yet it did not matter (of course it was a bit "easier" when they did) because they understood the above.
Advice to college students and recent grads:
Q: Help, I just discovered I hate programming and aren't any good at it. What do I do?
A: Go in to project management.
So programmers are run by people who are self-selected for mediocre skills or even outright hostility toward programming.