> an office is superior for productivity.

If I'm forced to commute to and work in an office 8 hours/day, 5 days/week, then I will be more miserable, lament my job, and thus be less productive towards my day job. I will feel more like a wage slave worker bee who has little autonomy over their work, having to regress to the humiliating dynamic of having to ask my boss for "permission" to work from home.

If I have the flexibility to WFH, not waste hours/day commuting, and live wherever makes me happiest (as opposed to the city my company happens to have an office in, which is probably a city like NYC or SF which may not be my preference), then I will be happier with my employment arrangement, more productive, and more loyal. Also I personally find it easier to accomplish deep work when working under my own conditions, as opposed to some open office with no privacy, constant chatter, and the drudgery of wage slaves who clearly don't want to be there but have no choice.

There are far too many variables at play to be able to make that kind of blanket statement with such confidence.

> Walking over and talking to someone is still far superior to any online communication.

The reason it isn't, for me, is that I'm usually answering questions about Why we do something, or How something is supposed to work. I'm usually giving a quick response, with a link to a design, or a code snippet, or a git commit, or a Code Review link, or something online. So I have to provide an online response whether you ask me in person or not.

If I'm in a 1:1, yea I totally agree I'd rather just talk about career growth and that stuff in person. But.... I mean now we are walking down the already-well-discussed path of Manager/IC and how these jobs are just different.

I don't know for sure, but I think this is the experience of many senior developers.

[edit] > I twiddled my thumbs waiting for the required SME to come online

I mean, this isn't an RTO discussion. If your SME is slacking off that is a different problem. Assuming best intentions, you would walk up to that SME's desk and there would already be 5 people waiting with a question because the SME can only talk to 1 person at a time.

I can slack with several different folks, so long as they can just wait a second while I type. The worst people (so sorry, there are some great TAMs) are the ones who page you as soon as they need something because they don't understand you are busy with a different project.

Yeah I guess the difference is that at my workplace there's a lot of tribal knowledge. Even when there is a design document it often hasn't been updated or requires nuance to interpret that's only really in one or two peoples' heads. Often the question/discussions is about engineering priorities, some discovery was made during coding that would alter the design, and half the time simply bringing up the topic with them for 5 minutes brings new information to bear that produces a superior result.

I try not to waste their time and develop my questions as much as possible on my own, but I'm not going to risk several hours of wasted work going down the wrong path when a 5 minute discussion with a SME would confirm something.

And it is an RTO discussion, because in the office they're available to some degree. WFH that availability drops off a cliff, and these are questions about the inner-workings of proprietary software, if the answers were on Google there would be legal action.

That again sounds like a problem with your company's knowledge sharing culture, not with the industry at large. Silos exist even when every butt is in an office seat, that's a well-known issue that far pre-dates pandemic remote work moves. The answer isn't "make everyone defenseless to random interrupts from other people" but "learn how to actually document things so you're not a company full of points of failure".

It seems like what you really want is for others to treat your emergencies as theirs. They're not, they're yours. If you can't wait 15 minutes for a ping back without being completely stuck, you aren't planning your work correctly. That's nobody else's fault.

So this is such a good example of why the office is actually a major liability. Being in person allows you to get away with creating a culture where documentation doesn't really exist and that tapping people on the shoulder is more efficient. Everyone will be better off if instead knowledge is shared efficiently.

A good analogue is the culture of "the smoke break"(see the Friend's episode The One Where Rachel Smokes for a funny example). It's easy to say "the best decisions are made during smoke breaks therefore smoking is necessary for efficient businesses." Clearly that's absurd. It's not the smoking that makes good decisions, nor does the ability to tap people on the shoulder make for good documentation. Depending upon an accidental formation of office culture is never going to go well in the long run.

> Being in person allows you to get away with creating a culture where documentation doesn't really exist

Viewed a different way, it allows a team to spend less time documenting and more time on other things. Increased documentation could be seen as one of the costs of working remotely.

Increased documentation is a cost you pay no matter what, you just pay it later in other formats when in-office - when the SME you treat as a guru finds a better-paying gig you have to scramble to rebuild their expertise from nothing or hope they are nice enough to do a thorough knowledge transfer. That kills productivity far worse and for longer than taking 30m at the end of your day updating some shared wiki.

All this to say, the trade-off you're proposing doesn't exist. You're arguing for paying twice instead of paying once - twice in the sense that you lose productivity now via random interrupts, and in the future when your siloed knowledge leaves you high and dry. Hope the "other things" you got done instead of documenting were worth it in that scenario.

Chat’s my favorite. Leaves a record so you don’t have to take notes or immediately record whatever you talked about so you don’t forget.

Screen sharing or pairing are better remote than in person.

In person’s good for a subset of types of meetings that some folks may find themselves in a lot. Brainstorming, product ideation, that kind of thing.

> In the office I just walk over to their desk. This is from the perspective of an engineer by the way, I'm not management.

One of the best things about WFH is this doesn’t happen.

You seem to have missed my point that all of those options require someone to actually be at their workstation and available. With WFH, half the time people aren't. In the office they are.

If WFH is going to be the future, then there should be uptime requirements with associated penalties so that someone is as available as they would be in the office. Software development is a team sport, you don't get to walk off the field whenever you feel like it just because you're home.

> require someone to actually be at their workstation and available.

No guarantee people will be at their office desk either to be frank. They could be hiding in the toilet, loitering at the water cooler, out on smoke break, getting coffee, ... etc.

Sure, but those activities typically won't last for hours. I've had to put tasks on hold because the SME wasn't available for several hours, without explanation or any communication whatsoever, to make a fundamental design decision in their area of expertise. And it wasn't a one-off or just one SME. People, at least my team, were getting constantly distracted. They essentially imposed their irregular and unpredictable schedules on the rest of us, despite being in positions of relative authority. I had to work nights I didn't plan to work because someone decided to respond at 8pm to a question asked at 1pm. That kind of delay would never happen at the office even if someone was slacking off.

Which goes back to the principle that WFH is more beneficial for senior devs than juniors. If you lack the knowledge or authority to work independently for extended periods, then instead of everyone being on the same objective schedule the junior people are forced onto whenever the senior people want to work, and more specifically when they want to be available to answer questions that only they can answer. It's like only being able to do homework during a professor's office hours.

> In the office they are.

lol what are they doing? Staring at the ceiling? What makes them always available in the office and not always available at home? How is this still a thing being discussed? Perhaps, at home, they are actually able to get some work done.

Again, this is my office. I'm sure there are some offices where people legitimately aren't doing much of anything at home. Statistically, that just has to be true. Maybe this is the wrong company?

> What makes them always available in the office and not always available at home?

I remember, way back in like 2012? 2015? Not sure when exactly, but we visited the office of a contractor we were working with, and they had a flatscreen on a stand in a public area with their lead developer on webcam, working from home. I don't know how everyone gets so utterly myopic that they forget you can just login to a videoconference (or IRC in the olden days) and just be available for these "quick chats" Bonus, if you need uninterrupted deep focus time, you can turn it off at your end.

And email, did we suddenly conveniently forget about that? LKML and countless other FLOSS mailing lists are proof positive that 100% remote can indeed build any scale of software.

I think you're the one who missed the point. Yes they are at their workstation, but they don't want to be available for constant interruption. To turn your story around. You don't get to interrupt my work whenever you feel like it just because we're at the office.

Maybe when we walk off the field and we're home, we can focus in peace on our work and be more productive?

Surely you’re capable of scheduling a meeting if you need to talk to someone or share a screen. Scheduling a meeting is far preferable to interrupting someone who is busy, in the office or not.

Seems like you're associating characteristics found with specifically bad employees as universally true for all WFH employees.

If the company needs you, you get to do it your way.

Also, software dev doesn't actually need a team, and, as the "key players" tend to go WFH more and more, the teams become worse and worse value.

This ends with the decline of "team managers" and "filler players".

Did you mean "team managers" in-metaphor? As in non-players, non-coach excess staff? Totally agreed if so, didn't want to rebut unless you actually meant 1st-line managers of teams of players (which I'd place as "coach" in this metaphor).

> Software development is a team sport, you don't get to walk off the field whenever you feel like it just because you're home.

Some of the most widely used software in the world (e.g. Linux) was developed without any such developer “uptime” requirements, with developers scattered around the planet, working on what they wanted, when they wanted.

This is obviously a very different thing than running a for-profit business. But there are countless examples of companies embracing the necessities of remote work to great success: GitHub, Automattic, Kinsta, Namecheap, Basecamp, GitLab, YNAB, Zapier, etc.

I am hearing a ton of issues with your company’s culture that have nothing to do with WFH. I genuinely hear your concerns as sincere, but, from the outside, all I see are warning lights. Anecdata, but I don’t know where my people, my boss, and any of my thousands of coworkers are, but I rarely wait long for a thoughtful reply, unless it’s “hey I am tied up but will get back to you EOD. Does that work?”

This a 1000%. Electronic communication means instant "paper trail" for free. Talking face to face is by its very nature ephemeral, OP seems to have missed that point.

> Walking over and talking to someone is still far superior to any online communication.

Hard disagree. That not only distracts that person, but also the six+ sitting around them. Most questions are fine as a chat message. We found people starting asking more questions to colleagues when Covid wfh started because they were afraid to earlier, communication actually improved.

The office is imo necessary for quite some things (onboarding new people, coaching juniors, people without a good place to work at hone, people who like the social aspect of the office) but walking up to someone for a quick question isn't it.

Being interrupted while I work because someone walked over to talk to me, when they could have sent a message to request to talk to me, is the second strongest reason I refuse to work in an office. The first is that I am not paid to commute, and commute time is non-trivial.

Having multiple tasks on the go is good. I did it even when I worked in an office; it seems common practice to have many active repos hot on disk.

> Walking over and talking to someone is still far superior to any online communication.

This is an example of a local maximum. This form of communication is better for the person doing the walking, but worse when looked at through most other frames:

- worse for the person being asked, who was interrupted

- worse for the people who weren't privy to that conversation but perhaps could have learned or contributed except that they were OOO that moment.

- worse for the organization because knowledge shared this way is ephemeral & siloed by definition

- ineffective in large organizations, where teams often span offices

And all that's before considering the "worse" that comes from having your staff spend 5 hours a week (average in the US) commuting to the office.

In effect, the person walking over is burdening the other party and the organization rather than spend a little more time to craft a question in writing. Weak written communication skills are at the root of why so many organizations are bad at remote work. If oral communications is all that works at your org, then remote is going to be really tough for you (and your inter-office communication is also probably terrible and you should put the whole company in a single building).

> - worse for the person being asked, who was interrupted

Generally, only in terms of personal output. I find that the organization benefits overall. Of course, this is problematic when performance/compensation is tied strictly to personal output, but that is a separate discussion

> - worse for the people who weren't privy to that conversation but perhaps could have learned or contributed except that they were OOO that moment.

Which is why working at the office is important.

> - worse for the organization because knowledge shared this way is ephemeral & siloed by definition

Somewhat true, but this is completely unavoidable. I know of zero cases where all institutional knowledge is recorded and not siloed in humans. Of course there are degrees.

> - ineffective in large organizations, where teams often span offices

This is why teams exist, why hierarchical organizations exist throughout history, and proper separation of work is crucial.

> Which is why working at the office is important.

I'm so glad you opened up this avenue for follow-up. Even when working at the office, people are not at their desks 100% of the time. People are still human beings and are out of the office for any number of legit reasons that have nothing to do with remote working: dental appointments, work travel, interviewing candidates, vacation, sick leave, etc. Regardless of how you organize your workplace or what mandates you set, people are going to be out all the time. So your knowledge sharing is among the best who are available right now, not necessarily the best sources.

> Somewhat true, but this is completely unavoidable. I know of zero cases where all institutional knowledge is recorded and not siloed in humans. Of course there are degrees.

The key again is the degrees. An organization can choose to bias toward using modern tools or not. "Modern tools" here also broadly includes "writing" because it provides so many advantages over synchronous conversations for knowledge sharing.

> This is why teams exist, why hierarchical organizations exist throughout history, and proper separation of work is crucial.

Here I call your attention to geography. There are any number of reasons for staff to be located where they are, and staff don't typically work on a single project over their careers. This means they will change teams from time to time. Combine these facts and you're going to end up with teams composed of people who live in different places, where it might not make sense to relocate staff (e.g. project only lasts a year).

An example might be: implementation consultants for a software company based in Chicago working onsite at a client in Germany. I suppose you could move the whole software company to Germany, but that only scales to one client. Which also highlights that people are often on multiple teams at the same time.

These concerns are most salient in the most hierarchical organizations, which tend to be the largest and have the largest geographic footprints.

> I find that the organization benefits overall.

Can you speak to how that is measured? I find these kinds of claims very difficult to support with any kind of actual data.

>> - worse for the people who weren't privy to that conversation but perhaps could have learned or contributed except that they were OOO that moment.

> Which is why working at the office is important.

What about the guy who just walked over to someone else, now both parties were absent in eachothers conversation.

This is actually a good example of: Falls flat if everyone starts doing it. So many times I walk up to someone in the office to ask him something, but he is doing the same so I can't find him.

You highlight another reason people are unavailable for synchronous conversations: they are in other meetings! The strength of writing is that it spans location as well as time. I find it odd that thousands of years after its creation, leaning on the written word can still be a competitive advantage.

You describe what sounds like slow moving places. When rapid reactions are required then even just being able to hear commotion a few rows of seats away can be quite valuable. Doesn't mean people cannot write questions well, but sometimes forcing attention now matters.

I'm not saying people shouldn't go to the office. I'm just saying there are a number of downsides to encouraging walking over to someone's desk to ask a question as a primary mode of knowledge exchange. These downsides have been enumerated at length in The Discourse. On the other side of the ledge is mostly "qualitative" arguments lacking in specifics.

The main thing is there are real trade-offs for the staff and organization, and those tend to get overlooked in favor of lifestyle or sunk RE cost arguments.

Do you think soldiers in battle should send well crafted emails instead of shouting or just acting on observations? Not sure things are just "qualitative"...

Let's get real, white collar tech work is nothing like battles. Even outages are not that dramatic.

Not like physical danger, bou can lose a lot of money in minutes in some places.

Absolutely wild to take a discussion about "Return to Office" and pivot to soldiers in battle. Solid B for trolling. Would be A-caliber except effective militaries write a lot and use written words as a key part of their command and control. (I will escalate your "soldiers in battle" and ask whether you think launch command on a nuclear submarine would launch nuclear missiles solely on the basis of a quick call?)

I know, actually, but any place relying very strongly on formalised communication over dynamic approaches tends to underperform anyway (and militaries relearn that every so often), so all this stuff will resolve itself over time anyway. The non-social crowd will lose out, just give it time.

Btw., depending on country and situation, nuclear missile submarines can actually have wide latitude on decisions.

>there are real trade-offs for the staff and organization, and those tend to get overlooked in favor of lifestyle or sunk RE cost arguments.

Employees are given some power and time back for the first time in decades after the opposite relationship ended up with being overworked, underpaid, and unappreciated as they are dropped at a moment's notice. So yea, I'm not surprised people aren't sympathizing for the companies who boasted about making record profits over the pandemic and suddenly they need to "downsize". burning decades of trust and goodwill has consequences.

Also, extremes rarely help. I'm not a manager, so even in office there are only people on my immediate team or a specific person or 2 I am working directly with on another team that I would do that with. When it is expected and we have the same tasks in mind. Even then it's better to get stuff in writing where possible so I'd rather send a message, even if it's just so I can give them 10-15 minutes to think over before I walk to them.

What urgent outages are you having that can't be circulated with a quick "hey is ___ down for you" in a shared org channel? Also, you still have to get up and walk over to the hubbub to find out what's happening, then walk back before digging into the issue - in that time you could have just read the slack message and been fixing.

You're inventing solutions to a problem that actually worsen the issue.

edit: also, automated alerting solves literally all your problems here - and that should be an investment even if you're at a company with 100% RTO. Your first line of alert will still be digital in that case. If you're waiting to hear panicked seatmates to solve issues you're sure not worried about immediacy.

It is always nice to know every failure mode or possible reason for intervention in advance, but not always the case.

I guess it depends how load-bearing you are for your organization/team. If I and everyone else worked in the office in 2 out of the last 3 jobs I had, I'd have an interruption at least every once an hour, and maybe a queue at times.

And yes, I should stop working with people that barely know how to work Git or Bash or the package manager for the language they use (even those I don't even use; if I indulge a request for help I often just become a glorified interface to Google). People just don't seem to value my time over their convenience a lot of times.

Platform engineers are for some reason always implicitly responsible for every engineering question that's not specific to the company's product.

> People just don't seem to value my time over their convenience a lot of times.

A good tactic is to ask them about rubber duck debugging when they come to you with a problem. If they don't know what that is, have them Google it. If they still come back after that, hopefully they actually have a problem worth more than one person's attention. Otherwise, they should find another job.

Quite often it's just a mix of laziness and knowing that I probably know better. I'm not sure if that's better than them building broken pipelines and custom kubernetes shenanigans that I have to catch later anyway.

For instance, we recently found some developer started using some very traefik-specific annotations and CRDs in our clusters (and they can only deploy manifests that don't pass through our renderer with a hack anyway). While we were migrating ingress controllers...

It just seems like there's quite a shortage of developers, so hiring standards aren't the greatest. Or maybe I should have a higher tier job. But I already have to work 4 days a week because while I get more done than my peers, I absolutely can not work for more than 1-2 hours a day if I get bored. And fixing people's shitty designs tends to be only somewhat engaging.

Walking over and talking? You can just jump on Zoom for a quick chat with a coworker. Text messages are low bandwidth, but Zoom calls are not.

Everyone on my team is available for a Zoom call when they're at their laptops, which is for the majority of the working day.

Just because your workplace didn't result in an ideal result doesn't mean that it's the norm.

In person I can walk by the person and see if they are available for a quick chat. Typing a novel is a pain when a water cooler chat will work. IF the person is busy then you dont shoulder tap.

I wonder how typing ability corresponds to wfh vs rto preference.

I can type at about conversion speed without trying too hard, so “typing a novel”, as you put it, isn’t a big deal to me.

>> Walking over and talking to someone is still far superior to any online communication.

Says who? And for whom?

I personally dislike being subject to disruptions on the whims of others. If they need something, they can ping me on Slack and I’ll respond when I feel like. If it’s truly urgent, they can call me over Slack (and I’ll answer if I feel like).

> So unless you have a job that requires minimal communication on a day-to-day basis, an office is superior for productivity.

Depends on the type of work you do. In my last job almost all communication was remote via video call (except a meeting with the CEO every two weeks). Commuting just to sit in front of a computer to talk to others made no sense -> I lost way too many hours of my life to driving needlessly around.

I'm all for meeting in person, if there is a clear benefit. Otherwise it's like with all the other bullshit, stop doing it!

> So unless you have a job that requires minimal communication on a day-to-day basis, an office is superior for productivity.

If that’s the standard then the commute time should be billed to the employer. That way they can weigh the full benefits of the productivity gain of in person communication, against the dramatic efficiency loss of 100 employees commuting to work.

If employers want the productivity benefit of in office communication, they can pay employees for their drive/walk/transit time.

"8 hours for work, 8 hours for rest, 8 hours for what we will" taken seriously

If you are on salary (as most devs are), I think it's pretty reasonable to include commute time in "workday" and not assume 8-9 hours in office on top of that.

I think that’s reasonable, but most management for most workers that are pushing RTO would not describe it as reasonable.

I’d encourage everyone to have the conversation with their manager so the expectations are consistent.

It sure doesn’t stop these same leaders in shipping jobs all over the planet for cheaper wages though where face to face contact isn’t happening without jumping on a plane.

> Walking over and talking to someone is still far superior to any online communication. So unless you have a job that requires minimal communication on a day-to-day basis, an office is superior for productivity.

Nope! All this does is break the concentration of the person you're interrupting. This is precisely the "bullshit" viewpoint called out in the headline.

I commonly find these sorts of views espoused by people who can't type very fast. For most people who aren't Really Good With Computers, it's much easier for them to speak to someone than it is to type to someone.

I make all of my staff that don't know how to touch type learn to type their first week. It's essential.

When I need to talk to someone, it's easy enough to message them requesting a quick video call and screen share to discuss "X" and if it's inconvenient, we just setup a time for later and it's done. The rest of the time, we've got a nice, quiet environment to work in. Ideal, if you ask me.

This is so weird to me. I’ve worked in wildly successful global companies to which I have contributed wildly successful work collaborating with people I have never once met in person, over years of working with them. I am, emphatically, not questioning your perspective, but I have literally never, over 3 decades, found a need for in person collaboration that could not be satisfied electronically. If anything, post-COVID norms such as Zoom/Teams replacing conference calls have improved my daily work immeasurably, even with much less shared space work.

I actually go to the office a fair amount but, far from a shared collaboration space, it’s largely a refuge for people who specifically want not to be bothered (by the dog, the kids, the neighbor’s leaf blower, etc.).

> Walking over and talking to someone is still far superior to any online communication.

It's a very good way to stop the person you're talking to from being productive too.

Have you seen some of the virtual reality from FB? I think it's getting close to the point where it's just as good as talking to someone in person, with the obvious benefit that not everyone has to be in the same physical office.

https://www.youtube.com/watch?v=MVYrJJNdrEg

Debatable. For example, software mediated interaction can transcribe, store, and share the results of conversations with little or no effort on the part of their participants.

You could of course wire up your workplace to do the same thing. But that would be creepy and intrusive.