Hacker News new | past | comments | ask | show | jobs | submit login
Building a walkie-talkie for remote work (pragli.com)
215 points by dsaffy on Jan 24, 2020 | hide | past | favorite | 164 comments

> quick, synchronous audio/video communication between teammates

..."kill it with fire" or other equivalent meme is the proper response to this, no matter how wonderfully it is executed.

Really, what everyone needs is more asynchronous communication and less attention draining and less interruptions! This can only help very small teams of extroverts with very-very-similar personalities work, it brings HELL for everyone else! And creates a homogenous bubble by locking people with different personalities out of the loop because they can't stand being in it.

We should put more time into bringing to the future ASYNCHRONOUS communication tools that WORK FINE, like email, comments in project-management-systems (eg. in Trello, Asana etc.)... we need more and better of these! (And without compromising them by adding push-notification and crap that makes them synchronous and attention grabbing.)

We also need more awareness of when to use async vs sync, and what tools are what.

Some people treat email and chat like sync communication, getting annoyed if there isn't a response within a short time. Ever have someone walk up or use chat to ask "did you get my email (that I just sent 5 minutes ago)?"

Async is great for initiating conversations or going back and forth a couple times. Go much more than that, and sync often (though not always) becomes easier and faster.

An example of an ideal flow to me:

Person A (12:34): hey, I'm trying to implement this new API call, but the parameters we came up with kind of conflict making it more complex than we thought

Person B (12:52): Hmm, maybe there's a way we can simplify those. Have time within the next hour to meet/conf/video chat on it?

A (12:55): just about to grab lunch, how about 1:30?

B (12:56): ok

A (1:26): alright, back, ready anytime

B (1:29): (initiates video chat)

(At end of call): ok, let's run this past rest the of the team and give us a day or so to think about it, and see if anyone has objections or we think of other implications. I'll send an email summarizing.

If anyone wants to dig into when to use sync vs. async, I strongly recommend learning about "grounding" when communicating:


The remote working place I'm at has a policy of starting with video calls, then voice if video isn't available, down to IM, then email. I'm not an extrovert at all, but I quite like it. We are a relatively small dev team - maybe 5/6 people within a company of 30 or so who we talk to less. It's generally faster to communicate like that, with the benefit that you get to know the people you work with a little.

Just clicking on an individual's face and being able to talk to them without acceptance does feel like a step too far to me though. That said, some years ago now, a colleague and I were saying it might be useful to have a teamspeak-style voice room where anyone in the team could join/leave at will but it meant you could push-to-talk to the room if you wanted to. The idea being that people could share their frustrations easily so the team could help them out rather than have them feeling alone and unable to hit the dial button to ask for help from a specific individual. And you could always leave.

We never got around to setting that up though.

I'm not an extrovert and I'd like this for people asking me questions.

Want to disturb me for 2 minutes and ruin my concentration losing me half an hour? If it's going to save you your afternoon - just do it. If you're probably going to send me a message on slack or call me there anyway I'd find this a lot less annoying. My goal should never be to get through the most work myself it's to move the company forwards. Having someone wait until tomorrow to get their job done because I haven't manually gone and checked my emails is a waste of company time.

Each to their own, and I don't think everyone would want to use something like this, but don't assume that it's only small teams and very similar extroverts that could benefit.

I'm with you on this. I don't understand the HN mentality of "100% performance at all working hours". Many employers do not need you to be highly performant all the time, but they do need you to be available all the time.

I'm also reading the Phoenix Project [1], and there's a worker Brent who knows every system well and can fix any issue. However, he is supposed to be working on the big company project (think like an ERP system). However, he can't get that work done because everyone and their brother keeps asking for a "quick" five minutes of his time, several times a day. All in all, he is not able to get his assigned tasks done because he's so busy with other interruptions.

If that happens then yes, talk to your direct report and come up with a better solution. But you shouldn't need to block out every interruption ever. There are times when it is warranted.

[1]: https://www.amazon.com/Phoenix-Project-DevOps-Helping-Busine...

Instant messaging allows you to set status and busy messages - alas the interface isn't organic as changing it would also disturb your flow.

Also the ability to control actions based upon what status you set is not granular or controllable enough. For example - if busy - only allow some people to send video mail message or IM for later perusal or if some people contact allow them thru if call more than twice, three times.....or if Director - instantly. Equally, to set a response time so if somebody in group A calls, make sure you call them back within an hour, two hours....10 mins etc. Details and control like that.

But back to changing that status, that is the important aspect that needs addressing as currently it in itself distracts from your workflow and having some setup that allows easier access. So if your in say vi, editing a system file - it sets you as busy, if your reading the news - your available, if you're reading email - your semi-busy and some people get thru, some don't.

So a system that detects when your busy, and sets your status would be the biggest improvement in work communications and in effect what we are reproducing is the personal assistant. An intelligence that screens your communications and handles what and when things get to interrupt you.

That is for me the area in most need of some revitalisation.

> Instant messaging allows you to set status and busy messages

That requires people to use and respect them. Half the folks on I use the corporate chat with are permanently "away", and "busy" appears to mean "walk over and bug me".

> If it's going to save you your afternoon - just do it.

That's the thing, will it actually be used like that? This feel so casual that it doesn't feel like it has to be used in urgent matter only.

I guess the important thing isn't about the tools but mostly about the training the user got around the tools. It's important to show what each one is used for and in which context. It's so easy to just believe something is more urgent than it is.

I think the key thing is that tools won't make people good at communicating. That needs to happen as either a life lessons or job training.

Something tooling can do though is help give cues. If I'm wearing headphones in an office frowning at the screen that's likely to change how people come up to me compared to if I'm clearly having a break. When remote those aren't around so you can't tell if I'm knee deep in diagrams and code or (say) hitting unsubscribe from as many emails as I can. Making those things clear to others is something tooling can improve on - but there's not much that can stop people contacting you if they're determined.

I completely agree that training and culture are an important piece of this. If someone pings someone every time they have a question, without searching for the answer on their own, it doesn’t matter what the medium is. It’s wasteful.

I think there's a place for synchronous communication as long as it's followed up by sufficient asynchronous communication such that people aren't left out.

Sometimes it's a lot faster/easier to be able to get on a 5-10 minute call to discuss something and be able to look at it together instead of sending messages back and forth for a week trying to get a point across / understand someone else's point.

I'll never tire of these responses to use case stories that essentially state:

"This is wrong because it doesn't work for me. Really, what everyone needs is more of what works for me."

These Use Cases are literally pushing what works for them.

There has to be a meaningful difference between "Here's something that works for us and might work for you" and "That's wrong, what everyone needs is this thing here"

Not sure if this startup is only doing 1-to-1 or it's 1-to-n, but if they start allowing 1-to-n, they'll likely run into problems.

A walkie-talkie system will likely fail for larger group sizes (e.g. >8) or more due to cross talk that will build up on a common channel. In the marine environment where ‘walkie-talkies’ are still in use, folks have worked around the problem by monitoring the common channel, 16, for interruption and then negotiate a switch to a less common channel to complete the conversation. In the long run, maritime folks will likely switch to TCP/IP to address, route and communicate more efficiently rather than the walkie-talkie system.

This sort of thing crops up in the office when a scrum master says to take it 'offline' in a standup, or when an open office gets big enough to have a din of conversations all going on at once.

Totally agree - I wrote a rebuttal to the concept of "virtual office" tools for distributed teams a few minutes ago:


This is a really thoughtful response! FWIW, we believe the future involves both sync and async tools. Friday seems like it could be helpful on the async side.

Curious, have you looked for solutions that work this way? (Disclosure: I am a researcher for a tool that works this way, but is aimed at a different market).

I have worked remotely for 10 years and raised VC funding for an idea very similar to this in the past – we failed to build a sustainable business, although the product did have a couple of hundred teams that swore by the product.

In hindsight I believe it's because we were following the startup mythology of "building a solution for the problem that you have" a little too closely. We failed to realize that this is really a problem that only small startup teams (usually a group of founders) have, as they need to keep in constant sync. It doesn't really scale past that. As soon as you have any sort of mixture of roles, personalities, or start to grow the team this is a communication method that just inherently doesn't scale well.

I do believe there is something here for small groups of founders/friends though, good luck!

Oh, you built Sqwiggle! A couple of friends and I used to work together remotely and we used it every day. It was awesome -- just the right balance between being unobtrusive yet easy to have a quick discussion when needed. We loved it so much, we even made a (much less-polished) clone after you shut down so we could continue with the same workflow.

I think many to one eventually becomes the real challenge. IOW how to get the message to the lowest ranking person able to deal with the issue depending on its content.

I think it's still fine. My experience with many-to-many conversation is that there is always someone whose computer is configured in such a way that it causes everyone's speech to echo and makes it difficult to understand. With a walkie-talkie approach, you bypass this problem entirely.

I think probably the commenter's startup failed because of competition from big hyped, well funded players like Slack. Maybe when the hype settles, they would have a chance. Timing is everything.

Only a handful of lucky few who are well connected don't have to worry about competitors, everyone else needs to assume a competitive landscape full of noise and misinformation.

I’m not sure that’s true regarding it being limited to small startups and founders.

I manage a team of over a dozen people at a large tech company with a few of them remote, and a pretty flexible work from home policy. We find that our standups can run long because it’s such an easy opportunity to chat face to face with remote people, but really you want a couple of conversations and doing them one at a time is kinda slow. It’d be nice to be able to just have a super quick multi party conversation without having to do a call. We’re even considering Mumble or Ventrilo.

I used Sqwiggle back in 2013 and while I appreciated the idea in theory, as an employee, it felt like an employee oversight tool.

Essentially, it gave the impression that my boss didn't trust that I was doing my work. This tool made it so someone could see if I was actually at the computer doing work.

It also drained my battery, so I always needed to have my computer plugged-in.

We used Sqwiggle back in the day until a picture of the naked wife of one of our coworkers got captured and stood there for over 5 minutes. That was the last day we used it.

Oh you're the company from that story? A true urban legend.

I'm sure it happened multiple times in multiple companies.

To add to that, it is also a problem that new remote teams have. When working with others who have been remote for years, we simply have learned how to use async communication for almost everything, and we just start up a slack call when we need to share a screen.

On the other hand, new remote teams are still trying to continue their habits from being in an office, so they want constant, easy communication. To me, this product sustains unwise habits more than it solves a problem.

But to each their own -- all teams are different, as are communications styles. I'm not going to say that products like this are bad ideas... just that the market is more narrow that people might believe.

We used Squiggle for a little while and really really wanted it to work for us. But the audio and video quality (and connection reliability) was just not nearly as good as FaceTime, and we were/are a two-man team, so we just switched back to FaceTime. I actually think Squiggle 2.0 but with the reliability and audo/video quality of FaceTime would be a different proposition. Audio feedback, background noise, broken connections, and laggy fuzzy video all make the experience suck. Details matter.

Indeed, AV is _hard_ – just look at how much people continue to complain about Google Hangouts despite all the resources they have. Quality of connection beats everything else which is probably why Zoom made it to IPO despite terrible UX

Thanks for sharing your experience Tom! I DM'd you on Twitter, would love to learn more.

It's almost like you and the other founders needed a better way to communicate...

My business partner and I have been working remotely with each other for the last 5 years, after a previous 7 years in a shared office.

To make it work, we bought dedicated iPads and put them on stands next to our computer monitors on our desks. We start a FaceTime call at the beginning of each day and literally leave it running all day. We mute when not actively talking, so we can listen to music (and not listen to the other's keyboard noises).

It is amazing. We always can see if the other one is present and/or busy on a phone call. If I want to talk, I just unmute myself and start talking. It is surprisingly similar to being in the same room as someone, without having to agree on the playlist.

Obviously, this particular setup doesn't scale to larger teams, and it does rely heavily on the very high quality and reliability of the audio and video of FaceTime. But I'd recommend it highly for those in the same situation. Made the cross-country move possible without destroying the business.

The dedicated device is key, so that it is always visible and doesn't take CPU or screen real estate on your actual computer. And I can't emphasize enough the importance of the high quality audio in FaceTime. We tried Squiggle back in the day and the audio/video were just not good enough (and the freeze frames were unflattering and distracting in a way that silent video is not, strangely). It would be nice to unmute both sides at once, but FaceTime is obviously not going to add that feature.

We thought about trying to make a product out of our insights/experiences but concluded that it would be a bit like re-implementing MS Word to fix the find dialog (if anyone remembers that issue back in the day).

That would require a very special, intimate relationship with someone to work. There is no way I could have a random coworker/boss' mug staring at me all day to my left. I wouldn't even want that with my wife/girlfriend.

Do you have a private office at work? It is like sitting in the same pod as a teammate.

But I agree you wouldn't want it with a wife and girlfriend. Could get awkward with them both there.

>Do you have a private office at work? It is like sitting in the same pod as a teammate.

And therein lies the problem: we should all have a private office wherever we work, preferably without crappy lighting and with suitable heat / cold temps and a comfortable chair and so forth. Instead we get open offices with the guy three feet away feasting away on his boogers all day and the guy next to us whose headphones leak noise everywhere.

This is such an issue for me that it is one of the top five reasons why I went freelance - so that I can work in enjoyable professional settings with good hardware.

It's really not like that at all. People aren't sitting right on top of you a foot away. If they are, I'm sorry.

> I wouldn't even want that with my wife/girlfriend.

My partner moved in full time last year, such that we now share an office in addition to the rest of the house. We're coping, but it is honestly more problematic than I expected as we both are on the phone part of the day and need focused work time at other points. A 1984-style telescreen of some remote colleague staring at me all day work drive me nuts, but more power to the parent poster if it works for them.

Aside from scaling problems, there are comfort/trust issues. I can only imagine doing this with certain people.

Which is really odd, considering I work in an office. Somehow the artificiality, the fact that it isn't a human puttering around near me, but a plane of glass doing magic makes a difference for me. But then I don't litter my home with surveillance speakers, either.

I felt "watched" for the first week, then it is just normal, like working in an office. The difference disappears.

If every broadband connection had its own public IP, that would be trivial to implement in a point to point way using any IP phone software out there (linphone for example) without any need for a central server. IPv6 once widely implemented might solve the problem, since new fixed IPv4 addresses are unobtanium for non corporate networks; I'm not holding my breath though.

> IPv6 once widely implemented might solve the problem, since new fixed IPv4 addresses are unobtanium for non corporate networks; I'm not holding my breath though.

I have this strange feeling that we (the general public) are, in some manner, being segregated away from full implementation if IPv6 for precisely this reason.

That the limitations of IPv4 have benefitted the centralization of our internet access; IPv6, due to its sheer size of address space, turns that space into a commodity that is easily shared. ISPs would (or should) cease to be "gateways" and instead become mere infrastructure.

Assuming efficient network bandwidth allocation, it would allow for easy p2p or "peering"-like arrangements, which isn't easy or as scalable with IPv4 today.

But of course, that isn't in their business interest.

I can't think of any reason why IPv6 isn't "fully deployed" today; I'm sure there's a lot of legacy hardware out there that doesn't support it, but I would think it would be a minority amount in the whole "grand scheme" of the internet. Certainly the "end points" - our workstations and phones and such - are all IPv6 capable today. There's no real good reason for the rest of it not to be - except for control over access.

Just my opinion, of course...

As a remote worker (and as someone who prefers remote work) I feel like this is attempting to recreate the feeling of stopping by someone's desk in an office. If that's what you want to recreate, I think this is an interesting approach, but I've always felt that remote workers do so specifically because they prefer not having in-office distractions. (Assuming those working remotely prefer to do so of course. I realize that's not always the case.)

> I've always felt that remote workers do so specifically because they prefer not having in-office distractions

Some, probably, but I miss that specific aspect about being in an office. I like that I can temperature and sound control my office, and be somewhere super sunny and cheap when the office is gloomy and expensive. I quite like this idea.

Depends on what you do.

My first job out of college was a database ops team that was distributed across 6 locations across the country. We basically lived and died over Nextel direct connect voice channels and AOL IM. It was a superpower.

We hope that statuses / our calendar integration provide the outlet for remote workers to have focus time. It's definitely a balance!

I've worked remote with a small team by having image of me on a monitor. I heard what was talked around the monitor and could push-to-talk comment back.

It didn't feel distracting the way receiving a phone call while being remote felt. It's clunky enough that people didn't come up unless they had to and I still could keep up with the verbal information that was passed along with the office gossip.

Some people get really bothered about this idea and probably if you are working on independent tasks you don't need to be so connected. For tight teamwork I found it to be the least bad option.

That sounds weird and awesome at the same time. Does the face move at all? Have you thought of using facerigging software?

I wonder what you think of virtual remote space with advanced face rigged avatars as the future?

0] https://facerig.com

I phrased it wrong. It was a stream from webcam to indicate I was there. I don't think it needs to be good quality or fast frame rate though. Maybe even better if it's really pixellated. Just something to remind people. If somebody forgot to turn on the tv I had to mute the chatter in my end as well, because people very easily start having too private conversations.

Avatar could be quite fun but maybe too uncanny as well.

We actually have a feature in Pragli that sort of does this... https://pragli.com/blog/live-avatars-with-faceapi-js/

Yeah I saw that but it's not as fluid yet. I am thinking of 3d faces with realistic facial texture and expressions.

Btw, nice product!

Oh yeah that would be awesome... I definitely think that's the future

You note in the article that the detection occasionally causes some lag in your overall application. Have you looked into a background worker for the task?

We actually discovered that the CPU increase was not caused by the detection, but rather by the CSS animation of the SVG moving around.

It turns out that the detection itself is very fast and lightweight, so there was no need to optimize (background workers would've been a good idea though).

We somewhat optimized the CSS animation, but it's still not where we'd like to be... if anyone on here knows how to improve CSS animation performance beyond dropping frame rate / shortening the animation, I'd love to know. Tried some obvious things (z index, will-change).

We used a similar setup to co-ordinate a project team split between New York and London. At both locations, at the end of the row of desks, we had a monitor at about head height on a stand running a google hangout. You could immediately see who was about and what kind of activity they were doing and then gather round the monitor for conversations etc.

This is very common in some industries to have always-on two-way screens connected to workers in another office. A company called Tandberg was the main maker of them. It’s owned by Cisco now but everyone still calls them Tandbergs.

Typically you would have one at the end of a row of desks to “extend” that row to another row physically located in another office and vice versa

I love Atlassian's implementation of this[1] where they installed "portals" around their offices, connected to a counterpart in another office - if you want to chat to someone you can walk over to the relevant portal and do so, but because they're always on and in public places they also facilitate the sort of spontaneous conversations you get when you bump into someone in the kitchen while making a cup of coffee.

[1] https://www.atlassian.com/blog/archives/developer-lives-save...

Genuinely curious: do people actually use these at Atlassian? In college we had a portal to another college that looked very similar, and it was never used.

At one point I worked at what had been Tandberg; and there were a couple of teams there with remote workers who were represented by a desk with Tandberg terminal on it with a always on session running all day.

That doesn't sound bad at all. Having a webcam pointed at my face all day and showing on some monitor would be very discomforting.

Set it up right and you could look all the way around the globe at the back of your own head :-)

Using always-on speaker phones (or equiv) during critical periods is pretty great.

I'd be keen to try misc variations of your telepresence strategy. From lowest to highest tech.

I'm also now intrigued by a walkie-talkie, CB radio (push to talk) style approach.

So it removes one of the biggest benefits of remote work, ability to concentrate?

We use Pragli, and I feel like it's the opposite. Rather than a Slack back and forth to see if someone is available, setting up a video conference, connecting, etc, you can pop in and say "hey, you have a minute?". They can say "no" or "can you come back in 10?". It's like stopping by someone's desk. If they really wanted to be heads down they can be in do not disturb mode. They can still show as online (so their presence can be "felt"), but have a status that says "Focusing" and block "calls". I think it's great.

>They can say "no" or "can you come back in 10?"

By then, the damage to concentration has already been made. Yes, you can set your status to "do not disturb". But you have to remember to do that.

Sure, it's still an interruption, but I don't think it's completely derailing. That's why I compared it to stopping by someone's desk. It still requires both parties to be responsible and aware.

Stopping by someone's desk is derailing, FWIW. I frequently find myself losing 30+ minutes before getting flow back from a 2 minute interruption.

One interruption an hour is enough to result in me not being able to get anything done for the whole day...

Actually, I'm starting to think I lost ability to concentrate because of that.

You've clearly not got a clue how a software engineer works if you think its not derailing.

I am not sure you really know how concentration works.

For example:


"We found that cellular phone notifications alone significantly disrupted performance on an attention-demanding task, even when participants did not directly interact with a mobile device during the task."

Imagine then what damage random people's voices asking you things while working would do to your concentration.

Did you read the Pragli article at all? It specifically addressss this.

I did. I see no mention of special do not disturb handling, instead high value placed on others being able to start voice communication without waiting for you to accept - the only thing that is opt-in is answering them, by the time your attention is already wrecked.

> no mention of special do not disturb handling

There's a section on setting manual statuses, including one specific example of "focussing - do not disturb". The animated screengrab below suggests that if someone wants to contact you with such a status set they will get a window informing them and the option to bother you anyway. Not sure how special you would want your handling, but something is definitely there. Though it isn't really clear to me if the example in the screengrab is based on a status or on a calendar entry.

I do agree that this isn't something I would want to work with, and as I wrote elsewhere, if my company had this I would set it to perpetual DND or just not run the software at all.

The resigned me would accept that DND mode because somehow it would be made hard to work without this software, possibly with HR involved. The naive me would think there's going to be something better or that people would actually respect that. The cynical experienced me wants a progressively higher level of pain delivered before you can call someone in DND mode.

They also mention that what finally made it feel nice is automatically setting DND mode when there is a calendar event (including any “focus time” blocks you create).

A big problem is that many people will ultimately ignore DND anyway.

The caller cannot ignore it, it will simply cause audio to not be accepted automatically.

From the HN guidelines: Please don't comment on whether someone read an article. "Did you even read the article? It mentions that" can be shortened to "The article mentions that."

Problem is most people are socially inclined to say yes. It’s just human nature. They are more willing to break the flow than to break the social norms.

I really feel like this part is something you could automate.

Give me a choice box of now, snooze bar, or several other options culminating in “take it to email, my hair is on fire.”

Before I retired, I avoided getting a cell phone because of this. Where I was working, I could see that my co-workers were using their cell phones as a company intercom. I didn't want any part of that.

You can also include a camera in your face and screenshare all the time...

I think the approach from some remote companies, which involved dedicated display with a camera that was _usually_ on (but could be _switched off_, just like closing doors to your office!) would be better than this "walkie-talkie"

Take a look at the section "Solving Distraction: Presence + Statuses + Calendar" in the article - I dive into our attempt to solve this problem.

I e always found it funny how HN seems to put remote work on such a pedestal... some people cant concentrate remotely. Kids are a huge distraction, as are many other comforts of the home.

It's funny how you make a generalisation and then are very specific. Not everyone decided to have children. I have designed my own living situation to be quiet and to minimize distractions. Even if your home is a very distracting place, working remotely doesn't have to mean working from home. You still get all the other benefits of openness and communication and you can rent your own small office with a door that closes.

I find it challenging to concentrate in an open floor plan office that is full of people. I find it absurd that it has become the status quo for software development.

Point taken.

The open office layout is a scam - pushed based on the build-out costs of new office space in a very rapidly expanding market in Silicon valley over the last two decades.

Companies needed lots of office space very fast. Companies also dies fast.

Building out office space is expensive and the push for open office layouts as a "culture" bit was just as much marketing as anything else...

It’s definitely possible to concentrate in an open office, it’s just an acquired/cultivated skill. Besides that, if someone on your team interrupts you and they’re not just being inconsiderate, they are actually enabling you to concentrate on the thing you need to be, which is this other person’s question.

But there's no way for the interruptor to know the priority of their issue relative to the your priorities.

This creates a major time management issue. If you have time blocked off for important, high value tasks, you can be easily distracted by someone else's lower level tasks.

Task switching, especially to focus on low level tasks, is highly inefficient in both the short and long term and should be avoided at most costs.

Sure, but isn’t that true in the remote world as well? I can’t imagine it’s very good form to just reject a call or ignore it when you’re busy doing something else, you have to provide a reply to the person about why you can’t talk, and then that’s just as good as having done the same thing to a person physically in your office.

I had a job where I telecommuted one day a week. On the days between, I helped people with problems, went to meetings, planned, etc. I could get a little code in every day, but any deep work would often have to be queued. I’d plot, scheme, research bug fixes and workarounds. Then on my remote day I would code like hell, all the stuff I’d planned out in the previous week.

On average, I got almost half of my code for the week done on that one day.

In my experience, ignoring someone stopping by your desk is very different from not answering the phone. A phone call doesn't have the same degree of immediacy nor the same expectations of a response

How is a phone call not the highest form of immediacy? Short of an actual pager event implying downtime and process, what could be higher priority to someone working out of office?

A voicemail. I frequently get calls, some of it may be a high priority, some of it not, no way to know, but I am not going to answer every call. If someone goes through the effort of leaving a voicemail I will interrupt my work and listen to it and decide whether or not it is more important than what I am currently doing.

I think it really depends on related context though. If I am working in an operational capacity then I may need to answer the phone every time it rings. If I am doing project work requiring a great degree of creative problem solving I am going to block off my calendar most of the day and prioritize a lack of distraction. My phone will be on "do not disturb" and I won't be looking at it. Impatience and interruption are the killers of creativity.

If you treat every inbound phone call as your highest/immediate priority, you have ceded all control of your priorities to other people. Unless you’re working in a call-center type of role, I can’t see how that could possibly be optimal.

I have a friend who uses an always on audio for his entirely remote team and swears by it. I think it might depend on personality types. I’m willing to admit it does work for some teams even if my own initial reaction is “no way!”

I LOVE having always on push to talk audio like mumble or discord. If you want to focus just log off. If you want to chat with the group jump in. Want to have a side convo without screen share? Pop to another channel.

Mumble (speex codec) is great quality and low latency. Just has a poor UI / UX.

I'm trying to convince my team to adopt a opt-in, voice only watercooler like that for some time, without success so far.

I makes a lot of sense to me, who have been doing this for over a decade with my friends.

Are you busy, don't join / mute yourself / go to the afk channel.

Are you waiting for a long build, just reading HN, between tasks? Join the channel and people can bond and talk about work/life.

So essentially a conference call that lasts all day, every day? I can see why your team did not approve.

I think that’s a gross oversimplification.

If you happen to check your replies I’m genuinely interested:

- do you work remote?

- if so, do you choose remote for solitude?

- if not- do you consider face to face interactions with your colleagues not as good a use of your time?

Again, genuinely interested because the full remote teams I’ve worked with that have voice comms at the ready have a lot less misunderstandings and the barrier to conversation is much lower than having to make a zoom / slack call.

I don't work remotely anymore. It's not that I think it's a bad idea. I think if a team isn't used to that level of communication, it makes for there to be some pushback.

Do you know a voice chat like this where I could get real-time information whether or not someone is there without having to join?

We use Hangouts and WhereBy (formerly appear.in) for some projects, but when you're in the "about to join" screen, the list of other people in the room isn't updated past initial load.

Yes, mumble has an API that can provide that info without having to join.

The UX isn't all that bad other than on mobile, which is one of the worst experiences imaginable. I use it pretty regularly and the comments about the quality and latency really bear out. There's a weird issue with the server that the longer it stays up, the longer authentication takes (up to 60s on our server now), but that's only a minor issue that's probably fixable trivially if we cared.

Hard to say if it works or not as there’s no measurements. How do we know if a new team member would perform better or worse under a different circumstance, e.g. no real time audio?

Making it very customizable is key. Different teams have different workflows.

I’m pretty sure we don’t actually need any remote specific tools. The last thing I want is random coworkers suddenly speaking in my headphones.

You can opt-out of the click to connect type of calls by setting the option in your status.

Opting out isn't free, though.

It's a social signal and one that, potentially, brings with it additional stress as you're actively opting out of something your colleagues/superiors believe is good enough an idea that they've implemented it.

This is a good point. If the manager is going to implement Pragli, they need to make it very clear that opting out is very OK and even encouraged... we need to do a better job of making this feel socially OK in the app.

This is the most eloquent and succinct summary of this concept I’ve read so far, well done.

> we

What a strange way to spell "I"

Maybe it's just me, but I prefer to not have voice conversations unless absolutely necessary. A service like this would be stressful, because I'd probably be the only one to stick to his guns and keep do not disturb mode on at all times.

I love the idea of taking well understood analog concepts and applying them to the digital remote workplace, but man, forced sync calls coming in on a whim without my control to ignore them sounds like my absolute worst nightmare.

If you haven't already, take a look at the section "Solving Distraction: Presence + Statuses + Calendar" in the article - I dive into our attempt to solve this problem.

Do you think our solution is insufficient? If so, any ideas on how we could further solve it?

I think having to clutter my work calendar with "I'm working" entries is a pretty clumsy way of signaling that I don't want people to randomly speak over Beethoven in my headphones. I my company introduced this, I would invariably "forget" to start the software at all, or would leave it on DND status permanently.

Personally I just don't see the benefit over a chat message saying "time for a quick call about <topic>? <zoom link>".

Its good until it becomes aware that you can use it to silently judge how productive you think other people are being, which then changes the way you work totally. Major benefiet of remote work is that you can work at a reasonable pace without being monitored constantly, and end up producing more because you are working at a pace thats more natural for creative work.

I think this is a cultural issue. You could say the same about JIRA. I've been on teams that have used it to closely track developer productivity and micromanage. In contrast, I've been on healthier teams that just use it how it works best for them.

Yeah slack statuses are already bad enough for micromanaging bosses

It would be easy to describe pragli as a conference call that never ends, but I'm not sure thats fair. Its one of various forms of telepresence, like a video wall or telepresence robot. If text messaging has taught us anything is that a chatroom is the right amount of distraction most of the time - until the VR workplace apps come (soon).

This seems like something I'd hate. I can at this point triage slack messages without interrupting my flow. Notifications can be trivially ignored or put on hold for 30 seconds or a minute. That's enough time for me to "save" my current mental context and text requires little brain power to triage so I don't lose my saved state for a non-urgent message. Voice is different, suddenly going from quiet to someone talking in my ear will grab my attention. I have to instantly drop my mental context and understand what they're saying. And it involves all the social parts of my brain which again kills my analytical context.

This is an even more egregious violation of someone’s attention than IM.

We tried to use mumble on one team with push-to-talk and it only lasted a few hours.

I grew up with IM and distributed teams online, and am very comfortable with it. It's my platonic ideal for remote/distributed communication... until expectations around response times are put into place.

If I can receive a message, respond to it four hours or a day later, and receive a phone call for anything more urgent than that, I'm happy to use IM platforms for collaboration.

> If I can receive a message, respond to it four hours or a day later, and receive a phone call for anything more urgent than that, I'm happy to use IM platforms for collaboration.

Isn't that exactly Slack with the "Do not disturb" mode on does or email?

You just described email.

No, because I can also communicate in real time over IM.

Additionally the social protocols are different. An email requires (some level of) planning and completeness. An IM can include just one question without reams of context, or it can involve a long detailed back-and-forth.

Questions over IM can be resolved in minutes or hours. Questions over email somehow morph into hideous monstrosities that waste days without anyone being satisfied with the answers.

And of course your mileage may vary, different organizations are different, etc...

Meh... synchronous communication is almost always worse than async communication in terms of efficiency, for example as the burden of thinking time of both sides has to get carried by both sides of the communication.

Seems like a Mumble server accomplishes the same thing. You can get general channels with an open mic or push-to-talk, and you can also do open mic or push-to-talk with individuals. Mumble is free and open source and runs on a toaster-quality cloud instance with CD quality audio and no latency.

Kudos for someone doing the IRC->Slack but for voice chat/Mumble.

Gave me a funny idea (its terrible I know) point a web cam at each remote worker (you) then give the boss a traditional remote control with all 20 "TV" channels. If the boss is looking an indicator light goes on and you get to have a one way conversation with the boss in listen only mode. (extra points for steampunk design TV/RC)

Never in my life have I wanted to use a walkie talkie in an office.

I've used walkie talkies a fair bit on construction sites and industrial plants, and there the value is obvious. Walkie talkies are a broadcast mechanism - everyone stays on the public channel, and you don't squelch unless you have something urgent to say.

On the face of it, it looks like Voice IM, with that I'm trying to figure out who needs this product or why its better than other options?

I don't understand where this fits in the hierarchy of human bandwidth (From Lowest to Highest):

Text Message

IRC/IM/Teams/et al


Phone Call/Conference Call

Video Conference

Face to Face meetings

From experience it's between irc and having an office mate. I've only done it with an always on videolink which is different from this.

The problem I see is that this seems to mean you need to learn radio dicipline which is hard, compared to having a video link constantly broadcasting.

Back in the 90s I once worked at an office that had a phone system with no ringer. You’d dial the extension and just start talking. Or you could “ring” them by saying something like “hey Jeff!”

As a new guy in that office, never having met “Jeff” before, this was really difficult for my introverted self to get used to. I was glad when I finished the trip and flew back to the home office where phones worked like phones.

This seems a bit like that, but with at least the ability to set some boundaries around who could start talking to you. If my current team adopted this I bet I’d spend my whole day in “focusing, don’t interrupt me” mode.

This sounds horrifying. Why make it even easier for people to disrupt you?

Oh... I was a little disappointed that this wasn't literally about walkie-talkies or a comparable technology, but rather some centralized internet-dependent computer software :(

Eh. Slack and other chat programs are just as fine, and would enable less distractions.

Don’t get me wrong, this is still a good idea. As a remote worker, I would not go for this solution. I already use slack, and google calendars.

IMO I don’t need more than that. And I think your solution is convenience oriented. Nothing wrong with that of course, people pay for convenience.

This seems more suitable for companies / employees that are semi-remote. I don’t think people who are fully remote would be for this.

We've been using Pragli for a few weeks. The team loves it! We used Sococo and Tandem in the past and this is the first presence app that is really being used.

Great job guys!

Thanks! :)

Both AT&T and Sprint (which acquired Nextel) still offer push to talk services. But they're not big sellers any more. What happened with that?

It's not the same, but I'm using Zello (IP walkie-talkie basically) with my wife and it is very convenient. For example when we are in a shop or have to quickly sync on something, especially if you have children. I imagine it could be quite helpful for remote work, but it depends on many factors, work time is one of them.

Synchronous communication is killing productivity for remote users.

Another completely different model is async communication using videos. https://apps.apple.com/us/app/devtendo/id1488872267?ls=1&mt=...

Not new. This existed a long time ago. It was called Nextel. Was hugely popular with contractors and in construction trades.

I really like the calendar feature, I feel like (if it’s used appropriately) it might encourage legitimate conversation.

While I don't think Walkie-Talkie metaphor works for remote teams in general due to becoming an open invite to interruption, it may work for teams that need to work in tight collaboration, like in rapid emergency response and battle management even when everyone is in the same building.

Why not just set up a conf bridge for something like that?

It would be interesting to have an independent study of the system, perhaps qualitative. Reading the other comments, not everyone is convinced by the idea and having scientific data may help. If the data hurts, that's life.

Surprised no one's mentioned Relay PTT devices. https://relaypro.com Looks like a lot of businesses are using them to replace radios now.

Quick note from the author: Pragli has a feature that is inspired by a walkie-talkie. Pragli itself is not just an online walkie-talkie. Check out our homepage if you want to learn more about Pragli (pragli.com)

May I suggest that "Building a walkie-talkie for remote work" might not be a good way to market it then?

As a remote worker for more than 10 years, the horror of the idea almost stopped me even clicking the link.

Is it just me or is the pragli.com website doing some heavy computations to everyone ?

I connected via Firefox on Linux and my CPU temperature skyrocketed to 81C almost immediately.

Note: Not this blog page, the main landing page itself.

Can confirm, eats up all 8 cores. Seems to be a script from firebasestorage.googleapis.com

Related (2 days ago) about pragli.com QA:


We actually built a "walkie-talkie mode" in Cyph (cyph.com) for a lot of the same reasons. Interesting to know that other people are thinking about this!

Culture is the biggest impediment to progress. Radio predates IRC by a 100 years. It's like most people DID NOT LEARN TO READ.

what's the benefit of this compared to a chat/slack request for a phone/videoconference call?

Not trying to be snarky, but did you read the article or did you skim it? The author explains the reasoning behind the walkie-talkie approach versus a phone call.

he does but I am not sold on the benefits. A slack message "can we talk, call me" would be async and non-distracting (just turn off slack alerts when you want to focus). Using an online call rather than on/off (like a walkie talkie) then seems much of a muchness.

Yes, and setting your status to "in a meeting", "away", "code sprint", etc. is helpful. These are things that could be automated, I'd consider that a better approach to the problem.

Let people know subtly you don't want to be disturbed unless it's important, otherwise you will be.

“Marco Polo” is a pretty bad ass app for this...

Reminds me of the book The Joy of UX

Congratulations, you just invented Skype for Business ;)

We have a lot of fun with this walkie-talkie


WebRTC Push to talk. It "just works" in most desktop browsers. No login, no registration. You can make private channels and there's several public ones. There's an iOS and Android client, too.

We really like the "push to talk" model. You can leave the client up and quickly say something to everyone else in your group. Especially good for non-critial, time sensitive things (i.e., "Y'all want to get lunch?") that are still important for a sense of "team"

Agreed. I like Walkie Talkies -- both real ones, and virtual ones. Push-To-Talk is a great way to communicate, and deserves to come back! It was very popular in the 70s, and remains useful among "hams" and people who work on movie sets, construction sites, etc.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact