Hacker News Reader

Hacker News

Comment by podperson | Hacker News Reader

Full thread

podperson

Don't fix email. Fix communication.

Why, on my iPhone, do I have:

1. An email app (which required a major update to unite mail boxes)

2. A "messages" app (which abstracts out two different message systems)

3. A phone app

4. A contacts app

5. Twitter

6. Facebook

7. Skype

What I want to do is (a) send messages to people (I don't care how), (b) check messages I've received (from anyone, using any method), (c) manage my messages (both incoming and outgoing), and -- as the writer of the article points out -- (d) manage my attachments.

On the iPhone (which is by no means the worst case) I might end up doing something stupid like looking up a contact, phone them, get sent to voicemail. Go back to the contact. Use a slightly different path to send an SMS. Discover it doesn't get sent. Switch to mail, and send a message.

Meanwhile the recipient gets a missed call, an empty voicemail, eventually gets the SMS, and then receives an email -- in three different apps on their iPhone.

Tiny incremental improvements to email will only nibble at the edges of the larger problem. Let me communicate with a unified UI and unified contacts.

View on HN

<snark>

In my kitchen I have:

1. A fridge

2. A coffee machine

3. A stove

4. A microwave

5. A blender

6. Mixing bowls

7. Measuring cups

All I want to do is feed myself. Why on earth do I need to have so many different things to do it?

</snark>

To...most people, having a separate "app" for things that do wildly different things is a good. Skype and email fill completely different roles to me, and I suspect they fill completely different roles to other people as well.

You might want to send a message to somebody and not care how, but I do care how.

If it's late, I might send my friend an email instead of an SMS because I know that the SMS will probably wake her up, and the email won't.

Being able to control this is a good thing.

Did you notice the descention of immediacy in your example of phone->sms->email? You went from the most demanding contact method "stop what you are doing and talk to me!" to the middle "stop what you are doing and read these 160 characters!" to the least "eventually look at this piece of text".

If you didn't care as much about your message, you probably would have done this in a different order.

If something major in my life happened (I'm having a baby! I'm going back to school! I got into YC! Somebody is buying one of my projects!), I would call my best friends to tell them, I wouldn't SMS them.

So would you.

And this is good.

> If it's late, I might send my friend an email instead of an SMS because I know that the SMS will probably wake her up, and the email won't.

That should be up to the other end to determine. I should not need to say `SMS wakes me up, email doesn't' when they are fundamentally the same thing. That's an implementation detail.

The right way to do this is for the other person to be able to tell the phone what they want, regardless of protocol: messages from my Nagios system at work, or family members, should awaken me. Anything else should be ignored until morning.

Sometimes I do want my friends to wake me up in the middle of the night:

"I'm drunk, I need a ride home."

"Can I stay at your house."

"$emergency"

"Hey, I really just need somebody to talk to..."

etc. etc.

I don't want to have to go through my entire contact list, individually setting people who can and can't wake me up at night. There is already a mechanism for this, it's called "human interaction". My friends know who can and can't wake me up; they know this because they're my friends.

By giving them my phone number, I'm trusting them to respect that, and only wake me up if they need to.

I've been wishing there was a layer between my phone and me that said, effectively, "You've reached David's phone - if this is an emergency or he's expecting your call, press 1; otherwise, press 2 to set up an appointment for a later call, or 3 leave a voicemail message." Where 2 would talk to my calendar and ideally their calendar and figure out a time that works for both of us without my involvement. Sure, people could just mash 1, but people always have an option to be rude when there's interpersonal interaction going on, and at least this way they have a way not to be rude while still reaching out to me by phone, and a way to actually get through to me in an emergency.

with gvoice, people can leave a voicemail that gets transcripted (vaguely) and texted to me. I glance at the text and decide if I want to call them back or not.

I love that capability, but it's not quite everything I want - I do want a way for people to loudly interrupt me in the case of a genuine emergency, but I want them to have to reaffirm that most of the time. The hands-off scheduling would be nice, too. But for what it is, it's rad, to be sure.

Mr. Number (for Android) comes somewhat close. You at least have the option of determining who can get through to you.

I'd prefer something more configurable as well -- priority list of who can reach me at any time, during daylight hours, when I'm generally open to talking, and never.

As well as a way for those with at least some level of trust (and, say, emergency services) to reach out to me directly.

Combination of Mr. Number and a pretty acerbic voicemail message ("Hi, you've reached me, don't leave a message, email me instead, if you're a good friend, feel free to text") low-techs this solution for me.

It's much easier and humane to accomplish this with multiple applets with different characteristics.

There are a lot of implicit social expectations built into different messaging systems.

I have friends who I regularly email, SMS, and IM with. We've never explicitly discussed this, but it's pretty much assumed that sending an SMS means you want a response ASAP, an IM expects a response within a few hours, and an email can wait a day or two. Even though all three are ostensibly messaging systems, there is a social understanding that different messaging mediums have different purposes; simply filtering based on sender doesn't convey enough contextual information to make a meaningful decision.

It's time to cook something! How about an omelette? You need to scramble the eggs first-- thankfully, there's a tunnel between your fridge and the scrambler. You press the button, and 4 eggs roll into the bowl.

While that scrambles, it's time to chop vegetables. You'd like to do onions, but the tube from the fridge to the chopper only fits carrots and asparagus. Not great for an omelette, but good enough.

Time to cook. Unfortunately, your chopper is out of date, and it doesn't work with your pan model. The scrambler, well, it's actually a mixer and was only meant for cakes, so it goes directly into the oven.

So you put special bowls in the fridge, suck the scrambled eggs and vegetables back through the tubes, take the bowl out, store it on dropbox, and give your pan a link.

God, this sucks.

I'm not sure how this really addresses anything in the parent post. Sure, interoperability sucks, but surely you're not suggesting that the solution is to glom everything together into a single monolithic app that does everything?

By all means, fix the fact that tools are painful when you want them to work together. Maybe unify a few of those communications tools.

> Let me communicate with a unified UI and unified contacts.

This was from the top comment. It is suggesting the solution is to glom everything together. I think the comment you are responding to is concurring with their parent comment by providing an illustration of the concerns raised.

I think a better example is how WebOS and Windows Phone handle communication. They treat communication like it's between two people, rather than between two applications. Text messages and Facebook messages are combined into one, so you can continue the conversation from anywhere. Facebook and Twitter posts are added together to show what the person is doing, rather than what accounts on disparate are doing. These aren't huge monolithic apps, they're more like feeds from services all tied to a single person. Both WebOS and Windows Phone allow you to combine multiple accounts for contacts together into one person.

Basically, it shouldn't matter what service someone is using. What really matters to people is people. If someone is at their computer, they might want to use Facebook messenger. If they're away, they'll want to use text. Barring that, they might rather use email. The point is, someone shouldn't have to open the Facebook app, find they're not online, send a text only to wait for a response and assume they do or do not have their phone on them, then send an email. If they can do it with one app, the app telling them the best way to contact someone, it makes communication much easier.

No one has gotten it right yet, but I feel WebOS and Windows Phone are closer than the traditional discrete applications.

No, there just needs to be interoperation layers like file systems, sockets, stdio, etc. In the kitchen metaphor, it's your hands.

Or you could mark the message as 'important' before sending it.

Alternatively, you could still choose email vs sms with a single toggle instead of having to use two very different apps for what is essentially the same thing.

<snark>

Like a lot of snark, it's not really that bright, but it's agressive and presenting itself as clever. Not all of those things deal with the same category of work.

> You might want to send a message to somebody and not care how, but I do care how.

But why do people have to remember how? Why can't I send off a message and have a computer figure out the preferred way to send the message. Remember, a big stakeholder in this preference is often the recipient. Also, most people don't really care about the how. They care about the result.

> If something major in my life happened..., I wouldn't call my best friends to tell them, I would SMS them.

You mean you would use the most appropriate form of communication, with the right degree of formality and immediacy. Right now that means picking from a slew of applications. The history of technology tells us that those applications will probably change, and that eventually people will be somewhat removed from the how, only expecting things to just work. (And that people who realize this can parlay that into tons of money.)

EDIT: Someday our insistence on picking an app to do communications will seem like driver's insistence on manually shifting gears.

If I send 50 SMS to someone who is charged 10c per message, I have cost them $5.00. So I should care whether I send them an SMS vs. an email.

Someone without a data plan doesn't have constant access to email. Maybe they only check it on a computer every few days. So if I want to send an instant message, SMS might be necessary.

Randomly sending an email vs. an SMS does not "just work". Nor does forcing each person to communicate to a computer some elaborate policy for how to reach them best.

> If I send 50 SMS to someone who is charged 10c per message, I have cost them $5.00. So I should care whether I send them an SMS vs. an email.

Yes. You're supporting my point.

> Someone without a data plan doesn't have constant access to email. Maybe they only check it on a computer every few days. So if I want to send an instant message, SMS might be necessary. > Randomly sending an email vs. an SMS does not "just work".

Again, yes, this is it exactly. Things right now are very far from it "just works." Things are horrendously complex, and it often sucks for one reason or another. Some people can manage all of this complexity in their heads, but that's certainly not true of everybody. What's needed is an integrated vertical.

> Nor does forcing each person to communicate to a computer some elaborate policy for how to reach them best.

True, that, though being able to set that just once might be better for some of the tech savvy than having to keep track of it all. That wouldn't fit everyone, though. What if someone could communicate their preferences to something like Siri after the fact, with a machine learning system adjusting preferences taking into account all of the communication stakeholders?

Paying to receive an SMS makes absolutely no sense.

Here in the UK, we pay to send them, not to receive them.

It's counted as one message each for the sender and receiver in the US. We are then billed simply by that number.

Conspiracy theories are bunk. That fact that US citizens stand for SMS double billing is the best proof that we're sheeple.

>the best proof that we're powerless.

More accurate. People need to communicate, and SMS is a good way to do that. Would we like it to be cheaper? Of course. There have been numerous lawsuit attempts, petitions, and boycotts. Doesn't matter. There's nothing we can do to change it. If everyone just stopped using it, they'd bundle it in with a package you had no choice but to buy. If you just abstained from a cell phone, you'd be in more trouble than you'd have been 20 years ago since payphones and emergency phones no longer exist.

It's not that people pay it and shut up about it, it's that people pay it because there's no other real option.

While I see a phone call as fundamentally different from an email, I don't see SMS as different. In fact, I see it as a really lame competing implementation of the same basic idea that deserves to die as soon as possible.

Funny, I see SMS as an improvement. Enforced brevity is an improvement IMO.

tl;dr will be the death of intelligent discourse.

There are lots of things I want to do with email that I can't do with SMS. Maybe half the emails I send could get across most of what they need to in 160 characters. Other recent messages included detailed tech support/troubleshooting, a list of equipment and design discussion with a partner about an application we're developing.

None of those would fit in 160 characters. There isn't even a way to losslessly compress them in to 160 characters. I can always write a short message when I only have a little bit to say though.

I also dislike twitter and like RSS. I'm evidently not in the majority.

The only functionality that SMS provides and email doesn't (in most countries) is network notification, which can be easily fixed. It would be relatively easy to replace SMS with email without users noticing the difference.

Software apps and physical appliances are very different, so your analogy is ridiculous.

When I email someone, software automatically routes it to the destination so I don't have to. Why should this not be abstracted across channels.

As for my trying to dictate how the recipient consumes incoming messages -- surely its more intelligent for each person to decide how they want to consume messages. I know some people who live in their SMS but ignore phone calls. Indeed the whole hierarchy of SMS / email / IM / phone / conference is very much in flux. Most people are probably more interested in who the sender is, not how they're sending.

But by your logic maybe I should use one app per person I talk to.

When I email someone, software automatically routes it to the destination so I don't have to. Why should this not be abstracted across channels.

There seems to be a whole world or metadata involving communications preferences which we've just barely started to scratch.

Sounds like your snarky self needs a matter replicator. I think the schematics are available in some book about Star Trek somewhere but you'll probably need to find a dilithium crystal to make it work.

> If it's late, I might send my friend an email instead of an SMS because I know that the SMS will probably wake her up, and the email won't.

> Being able to control this is a good thing.

those are great points, and a solution that can really replace the 18 methods we have of sending messages today would take that into account. Email today has this lame "importance" flag nobody uses but certainly, if we could get over the familiarity hump, having a unified kind of message where we can configure how it travels and how it alerts the receiver (not to mention, that the receiver would be able to route these various classes of messages in any way he/she sees fit) is not a huge technical issue.

Friend, what you need is a Thermomix... and $2k or whatever they cost! (http://www.thermomix.com.au for anyone who doesn't know what they are.)

I agree that having different forms of communication is useful. Phone/SMS is more personal, partly because I only give my number to those I want to contact me. I use my email for personal communication as well but I also use it to sign up for things like coupons from stores, etc. Some seem to be arguing that having so many options is difficult for the user, but I really don't know anyone who has trouble keeping them straight.

I realise that this may sound like I'm descending into fanboyism, but this is actually something that annoys me about iOS- it's far, far too app-centric.

Take Windows Phone for example: tap a contact in WP and you see their tweets, Facebook messages, your SMS and e-mail history with them. The messages app integrates SMS and Live chat seamlessly. Ideally this system would open up to allow any third party to plug into these hubs (and from the early signs of WP8, it will), but it's still already lightyears ahead of Apple's offering. Even Android, with it's system of intents and actions, provides more integration than iOS. I'm baffled as to why they've not tried to do anything about this.

Yeah, even webOS made a creditable stab in this direction. And that was what, three years ago?

Maemo on the N900 is pretty great too.

To expand on this, Maemo had a "Conversations" application that you could add accounts to. Out of the box I believe it did SMS, Skype and Google Talk. There were plugins available for third party services like Twitter, Facebook, MSN, Jabber, etc.

I love this feature and is one of the main reasons I use a windows phone.

As someone who has dealt with integrating and synchronizing data and accounts nearly my entire career, I can tell you this whole area is still in "there be dragons here" territory.

Here's a recent example of stuff Google did - merged my G+ circles into my contacts - ok, I can see how that'd be useful... but old or invalid email address for my mom and wife were substituted as first hits on the search when I entered the contacts in Gmail. Consequently, I've sent quite a few emails in the past few days where I'm getting bounces.

That's assuming you don't have to deal with improperly overwritten data, or time-sensitive data (ie, temporary details). Any degree of asynchronicity can create intolerable delays for example.

Abstractions can only get so far, especially for distributed synchronization across potentially adversarial vendors/systems.

If this stuff were easy, it would have been solved. That said, if Apple let us have the APIs, we could take a stab at solving it ourselves.

"2. A "messages" app (which abstracts out two different message systems)"

Is the abstraction a plus or a minus in your opinion? I'm asking because on one hand you're asking for a unified communication app, but on the other, the way you wrote that part, it sounds like a bad thing.

Taking this specific example, I think it shows that aggregating communication media is difficult and maybe not a good thing. In Messages, SMS and what is pretty much IM is mixed and I don't know which is which anymore.

I do find the idea of a unified inbox interesting and worthy of being explored but the thing is that each of these have different expectations from the users. If I send an SMS, it's typically because I expect a response fairly quickly: pretty similar to IM, except that I want to reach the person right now wherever they are. An email has a longer reply timeframe: if I don't hear back for a few days, it's usually fine. A tweet or Facebook status doesn't have the expectation to be read by everyone. (because it's one-to-many)

And I think the expectations go both ways: when someone receives an SMS vs. an email, they know the time to reply is different.

There are also different expectations in terms of half-life of the content. SMSs are sent and forgotten for the most part, while emails are archived for years. (though it's probably true that these archives are most likely not as useful/consulted as we'd think)

So I do care how I send my messages to people.

Even the messages app is almost too confusing, let alone combining everything else as well. As far as I understand it, if I send an SMS then anyone with a signal can get it; if I send an iMessage, then they need an internet connection or it won't go through -- is that right? If so, it's a pain to mix them up, as many areas do not get good 3G coverage (especially indoors) but excellent phone signal.

Surprisingly few of the people I text have iPhones (or smartphones at all), so I can't easily verify this.

Messages will send as an SMS if the iMessage isn't delivered in a reasonably short amount of time.

Interesting, but does that mean delivered all the way to the recipient, or just to the iMessage server? Is there good feedback? If I'm in a low-signal area, I need to know how long to leave my phone in that one spot that gets signal.

To the recipient, in my experience. And it tells you when this has occurred or if it is deferring to SMS for the send. The UX is very slick.

The problem is that you are trying to corral a massive set of different technologies and platforms into a unified 'god UI'.

This already exists to a large extent on my iPhone as far as sending outbound messages. I don't know about you, but when I want to email/call/sms/mms/videochat etc say, my father, I don't goto those apps. I just hit the 'home' button on my phone to bring up unified search, type 'dad' and get his contact page allowing me to instantly choose any one of those options to message him. Email, short message, video, phone call, -it's all right there.

If you wanted to take it to the next level, you'd want to be able to open up the API so you could toss in Fb, Twitter, or whatever other services you want on that page to choose from.

The unified UI is basically there already though.

It gets harder when we talk about incoming messages though, because we've got different senders pushing different types of messages to us on a vast array of different platforms and technologies. But isn't this what 'notification centers' are? A single point that unifies and corrals your collection of different incoming messages? And now Apple is merging their iOS notifications with OSX, syncing through the cloud. I agree it's not pretty yet, but we are getting closer, -like I said, I strictly use a unified message UI for sending outbound messages for most of my day.

I think this merging of central messaging UIs that are synced across operating systems will become more refined (so much room for improvement), but will definitely get there. But at the end of the day we're still funneling data from a wide range of different apps in order to present it as best we can keeping both form and function in mind.

I don't think I would want that. It sounds great in theory but in practice, there is a reason why I am on all those different channels. They reason is that they are different, even if that difference is subtle.

The One Big App To Solve All Problems is the wrong way to go, in my opinion. From a coding standpoint, it's a maintenance nightmare. From a UI standpoint, it's nearly impossible to create something that is both easy to use and comprehensively granular. From a user standpoint, it's confusing as hell when that "one thing" doesn't work like it's supposed to.

Yeah, you're right in the sense of why do we need all this stuff? The answer is: because no one uses just one thing to communicate.

On a slightly different note: if someone messages you and is expecting a response, do you really need to try three or four different methods to send that response? Shouldn't the expectation be that you would send a response back by the method that they made the request?

How will initiating a phone call through the same interface as writing an e-mail solve these problems?

You phone somebody and don't leave a message, so they get an empty voicemail. How can the interface on your device possibly solve that problem? (By the way, I have a $30 answering machine at home that's smart enough to not record empty messages, and for day-to-day use its interface is a button for "play" and a button for "delete". The interface is not the problem.)

Then you send an SMS and immediately turn around and send an e-mail as well. How will a user interface prevent the person from getting both the message and the e-mail?

I don't want to turn this into a (mobile) OS war. And my favorite is dead anyway..

That said, did you ever see the way WebOS handle[sd] this cases? The 'synergy' thingy was solving a lot of the friction points you're complaining about - and yet it obviously wasn't good enough to save the platform.

Some ideas:

- Sometimes we come up with solutions for problems that people don't have.

- Not well-marketed enough.

- Too complicated for people to grok. The abstractions get in the way of people trying to understand a feature. A new paradigm for communication would be hard to understand for most people and provide little benefit.

As a business owner, I can't think of anything worse than trying to maintain an archival system for a bloated platform that would combine all of that together.

I do however think that interoperability needs to be improved.

Sounds like this would be a great time for a rewrite of the UNIX Finger tool to solve a problem like this. Real-time client/address lookup and can be protected with some kind of pub/private key.

Right now, all of it is ending up in email anyway, except for SMS (maybe -- googlevoice can route your SMS to email). So you get the disadvantages of archiving all that crap (which actually turns out not to be that bad -- my gmail acct is only half full, and it even has wordpress database backups routed into it) without the advantages.

(Surely writing a rule to automatically throw away old voicemails, say, is pretty easy.)

Threadsy might have solved much of what you're looking for; unfortunately, they shut down.

http://news.cnet.com/8301-33200_3-57354466-290/threadsys-gon...

I would love to see someone else try to build this. Filtering options would need to be robust and flexible to combat potential noise and info overload. I could see marketers being an easy initial niche customer. General consumers, not so much. But I'd sure pay for something like this.

> Tiny incremental improvements to email will only nibble at the edges of the larger problem. Let me communicate with a unified UI and unified contacts.

Isn't that Google's main intent with integrating Google Voice onto phones? The recipient use case seems to work well under the Google apps where Google Voice shows your SMS and voicemails, a quick click would bring you to Gmail. The only issue I see here, is that I that I don't believe there is a cross-app indicator of new email (excluding extensions and external apps).

I'm not sure I understand what you mean by "a cross-app indicator of new email," but you can actually have voicemails and SMS forwarded to your email from Google Voice, creating a sort of "universal inbox."

Context matters in communication. Each of those platforms has vastly different expectations of the speed of the response. By unifying it all, what you've essentially done is made me wade through every message as if they actually had the same importance and priority to me. They don't.

Now certainly some things can be unified, but let's not pretend like every form of communication is equal just because in the end, a message is delivered to the intended recipient.

So in other words lets go back to big monolithic software suites controlled by one entity. I can't imagine a faster way to kill innovation and go back to the bad old days of things like the Mozilla Seamonkey Suite (its a browser, its an email program, its a html editor, and it sucks at all 3). Firefox stole that crown for a reason.

Seperate apps all competing for your attention/dollars is a feature not a bug.

That's actually what Maemo5/Harmattan had on my N900: all contacts and conversations across different media were nicely integrated through the Telepathy framework. It was probably the best app on the platform, although it took ages to develop and wasn't very easy to extend.

The unification of all my different messaging types (well, SMS, IM, twitter, skype etc.) is precisely why I'm still using an N900 - there's lots of other things it doesn't do brilliantly, but it does this specific thing way nicr than anything else I've used on a phone.

You must have felt the same way I did about the original Google Wave announcement.

So much potential squandered by a poor design team. I've always hoped some developers would turn the code into what it was meant to be.

What if I told you that a few smart people at http://vox.io are already on your case? :)

Windows phone does this. It collects everything into one place.

Google Wave tried to do all that and failed.

this is great point!

Reply on news.ycombinator.com