Hacker News Reader

Comment by davidhyde | Hacker News Reader

> "the biggest mistake I see engineers make is doing too much work on their own before looping in others"

> "For more senior engineers, it can happen because they like to work on their own and may be overconfident in finding solutions. It can also happen if the team culture is toxic and engineers fear getting criticism early in the design process."

Not saying the above statement is incorrect but here is an alternative explanation that is also viable: Senior Engineers are often hired into large large projects so that the company has the capability to address emergencies or modify the existing system accurately as it is often more difficult to work on a large complex code base than to build a new one from scratch. Those engineers sometimes never get to work on greenfield projects and "doing too much work on their own before looping in others" is a way to scratch this itch without the opportunity being taken away prematurely. It also offers chance to produce memorable work as nobody remembers the set of 3 point tasks you completed ten sprints ago.

View on HN

You are not wrong. I generally sell as a mercenary, and prefer it that way - I've been doing this longer than some of my managers have been alive. I'm paid just as well if they want to "pair program" or jira the whole process, but yeah, you hired me to fix a problem - if you are the problem, I get paid just the same.

Welcome to the real world, if your work is interesting, the clock might not turn on when I'm having fun. If your work is bullshot, I bill tighter than my lawyer

> I've been doing this longer than some of my managers have been alive.

In what organisation do 25 year olds manage 50 year olds?

Could a 30 year old manage a 60 year old? Certainly a 35 year old could manage a 57 year old?

If the 60 year old is wise and humble and 30 year old is open minded and not smug, it'd create a hell of a learning environment.

Definitely. Just do not micromanage and drag to endless useless meetings.

I have met several engineers that were very much senior to their managers but did not want to make the shift from IC to management.

I've known people who submitted working patches to the Linux kernel at 13, which I'd say is a reasonable start for "doing this". A 30 year old managing a 43 year old would not be strange at all.

Exactly, I started programming in AppleSoft BASIC when I was 11, picked up 6502 Assembly and Pascal before high school and added C, Rexx. 370 assembler and FORTRAN before I started college. I was 28 the first time I had a manager younger than me. I think the last time I had a manager older than me was when I had a job where my direct report was to the CTO.

Anecdote, but this was the approximate case on the team I worked on at Google (Manager in his mid 30s, team mostly in their 20s with one in his 50s). It happens. Actually, it was nice to see, as some evidence of non-ageism.

I hope to no longer need to work by the time I'm 50, but nobody knows what the future will bring, so I'd certainly like to have the option just in case!

A lot. Technical managers need not be experienced technical talents who have made a potentially incompatible shift to management.

It is better if the managers of technical people are technical themselves. Otherwise you wind up with the Dilbert "pointy-haired-boss" syndrome. The non-technical manager is extremely easy to bullshit, so it's better for the company in almost all ways.

An experienced manager is not easy to bullshit regardless. Sometimes a non-technical manager is better for many reasons. Having business domain experience can be just as valuable

I have encountered non-technical managers who were good, but in every case they would have been better still had they had some technical background.

I have encountered technical managers who were not good, for various reasons.

I think having a technical background is always a benefit, HOWEVER, it is neither necessary nor sufficient for being a good manager.

The same is true of software engineers. Every single one of them would be better if they had more of a non-technical background.

Yep; though enough time working for companies tends to instill some knowledge and understanding of the business side of things, and as long as understanding of the business correlates with their own influence it tends to work out okay.

Management is weird in that understanding of the tech side of things doesn't correlate with influence.

The problem with the "non-technical manager = easy to bullshit" idea is that it's basically the engineer's equivalent of the "engineer = assembly line cog who shouldn't be exposed to anything but their JIRA tickets" idea for bad management. It's what you find in poorly-run or excessively cheap organizations, but it's hardly an upper bound.

When you think you're fooling someone, you're really only fooling yourself.

It doesn't take a genius to know who is doing the work and who is looking for every excuse to be 'blocked'.

Managers simply know that they can't hire or fire and that calling someone on their bullshit would accomplish nothing, so they say nothing.

Managers are masters of soft power. Soft power is very hard for engineering/techie types to understand, which is why it works so well on them, because they don't even understand the game they're a part of :)

There is a lot of bullshit in this comment.

Yes. I hope we get rid of this cliché of the technical person who is not good at human interactions.

Most developers I know are good at human interactions, some even among the best.

I don't want to be put in that box, and I'm not willing to excuse someone bad at human interactions because they are technical.

Obviously some people are better at human interactions than others, and I'll be happy to adapt, but let's not imply causes.

Your anecdotal evidence and personal preference is irrelevant in a discussion about statistics.

Do you have statistics on the matter?

Yes. It may shock you, but there are highly educated, intelligent people on the internet and elsewhere.

There are also people who don't know how much they don't know.

It is rather unusual for someone to be called out on their bullshit. I’ve been working for 25 years and only seen it happen a couple times. Most people are conflict avoidant, so the perpetually “blocked” individuals are allowed to stay that way. I know people who’ve essentially done no real work for years. Who’s the fool here? Those of us picking up all the slack.

I have a decorated history of calling people on their bullshit. I have the 'you have been let go' and negative resume references to prove it.

The fool is the person not understanding the game they're playing.

Techies who do other people's work are perpetuating the game they despise - they are indeed fools.

Middle managers who perpetuate the game are smart, because it is their job to perpetuate the game. It is not their job to change the rules of the game - that's the job of the techies who can refuse to 'pick up the slack', let targets fail repeatedly and signal to upper management that the game isn't working, forcing them to change the game, which the middle managers will once again perpetuate, because that is their job.

I called someone out, and pushed back on someone who wanted me to do their work. This was a useless "scrum master" type who couldn't even update a spreadsheet for one of his weekly reports. I explained to him that that updating those spreadsheets was not an engineering responsibility, but I would be happy to provide him with input. I also complained about him to my manager. He was one of those guys who couldn't even copy-and-paste.

They just harassed some other person into doing it. He was a bit passive aggressive about it on some future calls: "Bob doesn't want to update the spreadsheet, so I will have Alice do it!" Anyway, nice guy, but didn't do any work at all, and he's gone now.

I prefer fairly non-technical engineering managers who stick to their lane. Their job should be to help manage the project schedule (in coordination with and guided by technical leads and product leadership), coordinate with other teams, coordinate with middle and upper management, and help their reports with whatever career development they need.

They should not be making technology decisions or specifying how the work gets done. They should trust their reports to not bullshit (and if that trust is broken, those reports should be fired). I think having a technical background can be helpful for these managers, but I don't think it's strictly necessary, and they should be doing essentially zero technical work as a part of their management job.

I've found managers who are like this to be incredibly useful and productive, and a pleasure to work with. Managers who want to get involved in technical decisions just get in the way and cause problems. Unfortunately a lot of newly-promoted former engineers can't let go of the technical work.

My first job out of university, I was promoted to manager in my mid 20s. One of the senior engineers reporting to me had kids older than me.

Now, they were PAID a lot more than me ...

The first thing they taught me was that it wasn't my job to do the work, because it didn't matter how smart I was, I didn't have almost 30 years experience.

I worked at a start up where the founders were 25 & 26 and most of their first 25 employees were their friends who were similar aged. In less than two years, there were almost 100 employees, plenty of which were in the 50s.

The company's last valuation has it worth over a billion dollars now and they have a few hundred employees. A good chunk of the C suite and VPs are still those same early employees who are now in their late 20s and early 30s. Some are managing former FAANG employees in their 50s.

Which one was it?

Not as extreme, but not far off. I believe I'm the youngest person on the team I manage.

I've spent a lot of time in startups - including way too many hours in my 20's working on side-hustles (wife was in med school, so it was kind of my thing to do). I've ended up in a situation where my technical skills are strong (but not the best on my team), but my business/startup knowledge is much better so I can help to ensure everyone is working towards the most impactful/valuable outcomes.

It's not unusual in the US military for a lieutenant or captain in their early 20s to manage many senior non-commissioned officers in their late 40s.

No mil. experience myself, but I'd say that while the officers may technically 'manage' the NCO's, that with few exceptions it is the NCO's that actually get things done.

Yeah the officers all move on relatively quickly; the NCOs may be on the same ship or class of ship their entire career.

A few years ago, I was a "team lead" at the age of 27 managing a team that included one developer over 50 at a big TV network company.

We're both at different companies nowadays, but I've been trying to get him started at my current company's team (which I am senior on, but not a lead) for a while now and he's considering it!

In a previous job I was team lead of a team that had a person in they're 40s and one in their 50s, I started as lead as a 23 year old and moved on when I was 30. There are many things I'd do differently now, but it was an effective and impactful mostly high functioning team. It does happen.

I'm a 25 year old managing a global team of 8 people. I'm self-conscious about my age both in my 1:1s with my reports and when talking with customers. That being said, it seems like things are going well. My team is doing great metrics and achievements wise and the company we work for tripled in size in under a year.

You know what I usually tell my managers. "Thank you for your service" :-)

Many organizations and it is a normal thing.

I used to work for a company in which employees in some departments would do 2 or 3 year "shifts" as managers. When your shift was up, you went back to a regular employee in the same department, and someone you had been managing took over. It was not uncommon to have a young manager, but not everyone was eligible, obviously.

I think it prevented the "us vs. them" mentality from creeping in, and I got the impression that most managers were eager to return to production work. There were dedicated managers higher up the ladder, of course.

How did that work out overall? I really like that idea, but I haven't heard of it before. People seem very motivated to stick to career tracks.

I’m also intrigued. I’ve seen this happen several times but always de facto: someone would become a manager as an experiment, try it for a few years, and then (usually) decide to stop.

They're still doing exceedingly well, so it must not have hurt them. I think it takes a small but stable company to pull it off. Maybe they have stopped doing it by now, I don't know.

These people cared about their work a lot, and they enjoyed doing it. Not sure which way the causation goes though...

Edit to add: the rotating managers I'm referring to were doing budgets and performance reviews, not creating business strategies.

I'm ... late 40s. Have been paid for working on software development in some capacity for... 28 years (first paid contract was 1993), and did hobby/amateur for several years before that.

I'm contracting with a couple different companies, and one of them... the others on the team range from ... 27-32. So... I've been doing professional/paid development work longer than most of them have been alive, certainly longer than any of them have been adults. The manager(s) I interact with - one just turned 31, and one is... I think 29 or 30.

The older you get the more common this may become.

> Welcome to the real world, if your work is interesting, the clock might not turn on when I'm having fun.

Ha! Yeah, I worked on a really cool project for a charity a few years ago: "How many hours is this going to cost us this week, you were working like a warrior?" "35"

> If your work is bullshot, I bill tighter than my lawyer

Lawyers must hate their work.

I believe it has one of the lowest job satisfaction of any white collar jobs. I've wondered that is folks didnt have the massive student loans to pay off early on if they would stick with it.

Many law firms are incredibly well paid sweatshops.

Never mind when law students graduate and go to work for the government, for a salary less than a year's tuition...

Don't they get student loan forgiveness? Or just all the other kinds of benefits.

I can't speak to the general case, but I know that my brother didn't. As an ADA (living in an expensive US city) his salary is less than a first-year public school teacher with a bachelor's degree.

> Lawyers must hate their work.

A friend of mine is a criminal defence lawyer. I get the impression he really enjoys his work – he gets to meet a lot of people he never normally would (bikie gang members, terrorists, murderers, drug dealers, drug addicts, etc) – and he feels safe in doing so (he tells me that defendants trying to harm their own lawyers is quite rare, rare enough that he isn't worried about it).

Once, at a work function (previous employer), I met one of the lawyers from the contracts department. He was telling me how he used to live in a rural area doing agricultural real estate transactions, now he had moved to the big city to do in-house contracts review for a multinational software company. He was a "top performer" (indeed, this was a function to reward people who'd been nominated as "top performers" by their management)–but he didn't give me the impression he really loved what he did, more that he was just doing it to support his family.

I think criminal defense is a great service to humanity. It's hard to estimate how many are wrongfully accused, but surely there are many. Further, things like the plea bargain system or parole regularly lead innocent people to proclaim their guilt. It's very messed up and must be a terrible trauma for some. Lastly, I think even guilty people deserve humane treatment and perhaps forgiveness.

> and perhaps forgiveness.

I would agree here, on the condition of rehabilitation.

Obviously our "justice system", isn't.

Edit: Speaking of the US above, no experience or knowledge about other countries systems. It appears I've found a new gap in my knowledge, anyone have a good intro to how courts work in their country?

My lawyer friend and I are both Australians, and he works as a criminal lawyer in the Australian legal system.

The US inherited the basics of its legal system from that of England. Australia did too, along with many other countries. So at a very high level, the basics are the same. But, there has been a lot of divergent evolution, so as we drill down into the details lots of differences come up.

I think one huge difference is not really legal but social – Australia has always been a less violent society, with less violent crime and less social conflict than the US has, which reduces political pressure for punitiveness in the legal system. On a per capita basis, the US homicide rate is 5 times that of Australia, and while these numbers go up and down, I think it has been consistently significantly higher than Australia's, for many decades.

Can you shed some light on rehabilitation in Australian prisons?

I don’t really know. My uneducated impression is that Australia’s prisons are harsher than those of many Western/Northern European countries but not as bad as those of the US-but I don’t know of any hard data on the issue. Rehabilitation is officially an objective and of course the prison system makes some attempts but we can always question if they do enough.

In Germany, they use the inquisitorial system. There is no jury, only judges. In smaller matters, it's one professional judge and two lay judges. In bigger matters it's 3 professional judges and five lay judges. The professional judges hold the ultimate decision, and the lay judges advise (but they do have a lot of influence). And of course one can appeal.

There is no such thing as plea bargaining (which is considered a perversion of justice) or bail (which is considered to be jailing the poor). People are rarely jailed pending trial (unless their place of residence cannot be established, or they're considered a flight risk - a very high bar to pass). There's no "perp walk" or handcuffs or any of that humiliation stuff (unless you actually ARE being a pain). After the verdict, you're usually given a month or two to tie up your affairs outside before being ordered by mail to report to prison.

The relationship between lawyers is one of cooperation (albeit with the agenda of pushing towards their point of view). The purpose of the court proceedings is to get to the truth of the matter. There are no theatrics, no grand speeches, no "chewbacca" defense. Withholding evidence or deliberately hampering the other side is severely frowned upon. "Winning" the case is not the point; representing your side well is. Lawyers have no election aspirations, and thus no public to impress with their prosecutorial prowess. One cannot become a judge without years of training and a degree (there's no such thing as elected judges).

The result is a very calm, organized court proceeding, because everyone involved knows the law and is committed to it. You can't bamboozle with doublespeak, shady arguments or unproven methods because there's no "man on the street" juror to deceive; only the judges, who are trained to know better.

Anyone can sit in on a proceeding (I've done so). There's very little ceremony, and often it's just a bunch of people sitting around conference tables in a small room.

I have a theory that lawyers in many ways are not unlike a security engineer:

you get the satisfaction of exploiting the system, bending the rules to your advantage, making the impossible work. and when you’re good at your job, you get paid well ( lawyer by many more times of course )

case n point: defending OJ

Lawyers aren't tracking & billing in 6 or 10 minute increments to punish their clients, they do it so you only pay for the time they actually spent on your account. It's beneficial that, if they get a 5 minute phone call while preparing a document for you, you don't get billed for the time they were on the phone.

Yeah, it's wicked beneficial that you're paying $40 or even $80 just to say hello and ask how their weekend was. I understand that from the other side, talking to you is working and they wouldn't be doing it if they weren't getting paid. But when one's rate is in consultant territory (as opposed to lower contractor rates based on bulk time) then that type of overhead should already be built in. And sure billing increments are theoretically orthogonal to billing for overhead, but it's galling to see "1.1 hours" knowing that 6 minutes of that was overhead that didn't get rounded down.

OK... so don't call someone who bills you for their time and make smalltalk. That seems pretty self-evident.

If I start a meeting with a consulting software engineer and spend the first 5 minutes making small talk, that's fine, but my company is gonna pay for that time. The same thing is true for lawyers.

Yes, that's the obvious conclusion. But surely you can see how clients needing to deliberately refrain from standard pleasantries doesn't contribute to lawyers having a good reputation.

I like the cut of your jib

Haha I like everything you said except the "pair program" bit.

From my point of view as both a programmer and an employer, this is the cleanest way to make sure knowledge transfers between short-term and long-term team members.

Why the disdain? (If I interpreted correctly.)

Pairing is only helpful when you are new and learning from someone more experienced.

However, it may be a time-sink for the more experienced person, who could be using their limited time on more critical things.

That's why newbies get assigned a mentor, with whom they can pair freely.

But pairing rigidly with the whole team just sucks and is incredibly unproductive.

I don't think it's true that pairing is only helpful for new people learning from someone more experienced.

But I also think it will be difficult to convince you otherwise if that's what you believe.

> But I also think it will be difficult to convince you otherwise if that's what you believe.

This is true. I have been on both sides of the pairing equation. I am currently on a team that pairs too often and I am annoyed on a daily basis because I could have been utilizing that same time for actual work.

So yes, it's going to be difficult to convince me but feel free to take a stab.

Okay, there are three other common scenarios where I have observed pairing to be immensely valuable:

1. You are in the middle of a death march and pairing with a team mate gives both of you the moral fortitude to keep going (and keep your programs relatively correct and relatively secure).

2. You are working on a tedious but difficult task and working with a team mate makes it so that you are less likely to make mistakes despite the tedium (as well as gives you the moral fortitude to continue with the task).

3. You are working on a difficult problem and you have a team mate that you can truly collaborate with to find solutions neither of you would have found on your own. (The Jeff Dean + Sanjay Ghemawat collaboration at Google is the most famous example of this.)

In any of these cases, it is important to pair with someone compatible with you. And you can build compatibility over time.

All the above examples do not describe my day-to-day experience. Are you literally on a death march 5 days a week, every week? Are you literally on a tedious but difficult task 5 days a week, every week? Are you literally working on such a difficult problem that it needs true collaborators 5 days a week, every week?

I specifically want to talk about - The Jeff Dean + Sanjay Ghemawat collaboration at Google is the most famous example of this.

Do you think they really produced all they did by pairing together? In my experience, the best things were produced by independent thinking, collaborating on experiments and trusting each others decisions. Not by actual pairing on the screen together.

Who said you have to pair five days a week every week? That doesn't sound reasonable. Very few things should be taken to an extreme like that.

For Dean-Ghemawat, I think they produced many great things by pairing on the same screen together. They also obviously did good work independently of each other.

I've had similar experiences.

I vocalize a lot (in my head) during the development process, and I've noticed that when I have to engage with another person during this time I start vocalizing audibly, which is where the problems start. The other person gets confused because I'm speaking the snippets and half-ideas that are passing through my mind at the moment (and the internal jargon I've developed over the decades), and it degenerates into long sequences of explaining my every move, most of which would have been quickly dismissed internally had I not had to stop to explain them.

Or they misunderstand the path I'm taking for something else that won't work and interrupt me, and then I have to stop to explain first that it's not what they think (which is hard), and then explain the path I'm actually going down (harder), and why I think it's a good idea (hardest). It's like having a backseat driver, and just as infuriating.

Every interruption shatters parts of the intricate glass tendrils that comprise the complex model in my mind, and then I have to rebuild them again before I can continue thinking. The creative process gets completely disrupted and there's zero chance of flow. This is fine if I'm teaching someone since the material is well known and there's no creativity required. But during actual work it's a productivity black hole. Death by a thousand cuts.

But imagine if you and that other person could develop a common language and share tendrils. :)

That would be akin to jazz musicians being in sync, which is pretty rare.

> It also offers chance to produce memorable work as nobody remembers the set of 3 point tasks you completed ten sprints ago.

This is one of the more toxic ones. To get past senior, you often need to be seen to do Big Memorable Things. It sometimes leads to perverse incentives.

I mean it's not that toxic, if you are a good worker bee your manager will usually notice and be happy with your performance. Then from time-to-time you branch off to do something more high-risk to add to your promo doc.

I think a lot of people have weaker communication skills than execution skills. So they could loop in everyone early on, but their idea might get killed off because they failed to justify it properly. If instead they leverage their execution skills and make an MVP that will speak for itself, then they bypass that issue.

Maybe toxic is the wrong word. Perverse maybe. But I don’t think that being a good worker bee often gets people past the senior level, at least from what I’ve seen. But also getting past senior is rarer, so maybe what I’ve seen isn’t representative.

>"if you are a good worker bee your manager will usually notice and be happy with your performance"

And other than some bonus never promote you. If you have capability to be anything above that "worker bee" say / ask exactly what you want. If not look for another job. While this SCRUM / Agile bullshit is wide spread and even works in some specific circumstances there are enough companies that are not hung up on moving pins on dashboard and where one can really grow.

Work for yourself. Work with the manager, not for manager.

> Then from time-to-time you branch off to do something more high-risk to add to your promo doc.

I agree, that's exactly what I wrote.

> If instead they leverage their execution skills and make an MVP that will speak for itself, then they bypass that issue.

Hell yes. Forgiveness beats permission every single time.

Exeactly. There is too much emphasis put on rewarding those who make big, disruptive change and large, solo wins. Someone making a lot of little wins are just as valuable yet usually don't any great accolates for it. Not to mention, putting one person on a large-ish project is a great way to create silos.

I think it's management and their "agile" philosophy that's toxic here, intentionally depriving engineers of any sense of ownership, autonomy, or vision-fulfillment over their work by constantly bouncing them around across small disjoint tasks.

I am incredibly fortunate to work in a place that values ownership, both explicitly and in practice. Given how widespread the opposing value system is, I'm afraid to ever leave.

because programmers must be replaceable parts for the industry to keep going at status quo

We have plenty of turnover and it works out fine. In fact turnover of the senior engineers who are service/project owners usually creates the opening for one of the more junior contributors to demonstrate L+1 competency by taking over.

“the status quo is fine”

Of course the status quo is fine, you're one of the highest paid people in one of the cushiest occupations in all of human history.

To me moving beyond senior means you enable the entire team to contribute together accomplish Big Memorable Things.

This can mean a ton of things, which cannot all be served by a single senior+ engineer:

* mentorship

* seeking out, establishing, and evangelizing best practices, and not just coding best practices: architecture, documentation, testing, ci/cd etc

* high-level architecture knowledge and experience

* evaluating technology choices: tooling, databases, orchestration platforms, etc etc

* assisting management and product with scoping and prioritizing work

* the ability to put your head down and crank out a solution to something in code simply because it needs done and you can do it better and/or faster than others

* laying the framework of a greenfield project, maybe sketching out the codebase or POC for juniors to take and run with

* ...and so on and so forth.

A single person may be able to contribute all these things to a team over a time frame of multiple years, but in a 3 or 6 month time frame, most mortal engineers could only contribute two or three.

If every senior only does internal assistant work, which senior does actual building of complex services?

What is more important? Building better services or being a library writer?

Every point you describe is exactly perverse incentives.

"Write great code to build stuff" is still a core part of the job.

But it doesn't really enable anyone else on the team, other than providing a model of well-structured code to follow - and most people's code is really not as great as they think it is.

I'd rather have a team of average coders who do things to enable and better each other and the team over a team reliant on a great coder or two who can't or refuses to enable anyone else.

It is not even toxic. Modern development is cooperative to absurd level all too often. You don't have any ability to make independent decisions or autonomy pretty much any time. Literally everything is result of negotiation and compromise and what not.

Taking chance to do a bit of focused work and actually do it in that environment is not toxic. It is not like you have run of for months and refused to communicate. It is just ... getting slight bit of rest.

Yes, I think this is undoubtedly true. I didn't really learn until I advanced my career that, in some respects, being a junior or mid-level developer is actually better. Being a senior can be a slog, and while I got paid little in earlier positions, I actually got to do more interesting things, screw up more, and build things from scratch.

The other thing that can encourage doing too much work before looping in others is the culture of the team. You are lucky if you can amass a team of people who all have the same attitude of talking to each other frequently and handling a certain level of interruption. I've worked on such teams and miss them. Although pretty much everyone at any company you work at will say "ask questions, don't hesitate, blah blah blah", there are teams where people go weeks not talking or sharing work, and everyone is always too busy to get interrupted. If work gets assigned in large chunks, as can happen for seniors in particular, it can also just not make sense to loop in someone temporarily if more time needs to be taken for the other person to catch up.

In my case, I'm the senior engineer and the rest of the team are juniors. There's generally not much I can ask them that they'd know the answer for. On the other hand, looping them in is a good learning experience for them even if they can't help me solve the issue. But due to time constraints that isn't always possible.

You nailed it. Senior engineers build things on their own because their company can’t afford to hire enough advanced enough people who can learn the system and take over the work without being trained for 6-12 months, and often there is no budget for this kind of ”distraction” from the core moneymaking path.

A good point, but the reason I have (and fight) this problem is much simpler:

I like writing code. I dislike talking to people.

Or, you work at an organization which gates promotion on leading projects, so everyone has to lead things.

Or your organization relies on "consensus" or "influence" to get any project done, and one of the groups (which don't share managers back to the ceo) doesn't want to do the project even though the others think it is important. So do you spend months trying to convince them while doing nothing promotion-worthy in your own?

i see the point here.

i do think giving space for hard work is important. i still more often see the opposite problem of too much silo'd work leads to wasted effort, but for some really hard problems, having a bunch of space to think them through is beneficial.

Reply on news.ycombinator.com