Hacker News new | past | comments | ask | show | jobs | submit login
How we built Flow (medium.com)
197 points by katm on Nov 25, 2014 | hide | past | web | favorite | 46 comments

I love any and all attempts to develop, improve, explore, the interface between humans and computers. I think it is one of the areas where not nearly enough is being done ... as a player of the continuum fingerboard, I quickly realized that a transformation of interface can have a more dramatic and meaningful impact on sound than the actual synthesis engine.

Roger Linn used to have a great demo on his page showing him playing a prototype Linnstrument using only a sawtooth wave with a lowpass filter, and it sounded more expressive than many top of the line physical modeling synths, because the dimensions of control were so impactful.

That said, the Photoshop example you give seems a little bit of a stretch. The Wacom tablet is a wonderful piece of human interface for Photoshop that addresses many of your concerns, and many other concerns, and has existed for years. For example, some Wacom pens have a ring on them that you can use to change brush sizes.

Wacom Tablet: http://blog.phaseone.com/work-faster-by-customizing-your-wac...

Continuum Fingerboard: http://www.hakenaudio.com/Continuum/

Linnstrument: http://www.rogerlinndesign.com/linnstrument.html

If you want programmable high-resolution knobs, this is an area that the music industry has had covered quite well for a long time. A quick Googling for "midi controller knob" brings up:

Livid Alias 8

Livid DS 1

Nektar Panorama P1

MIDI Fighter Twister

Behringer B-Control Rotary BCR2000

Akai LPD8

The last one is about seventy bucks and gives you eight independent knobs and eight pads.

Of course, there's value in minimalism and a good UX for configuring/programming the knob(s), but there's also something to be said for just trying to solve the problem yourself based on what's currently available. If a $70 MIDI controller makes your Photoshop experience noticeably better, great. If not, a single knob might not end up being worth it either, regardless of how well-made it is.

Also such controls exist for the CAD/3D market for a long time.

For example 3dconnexion has a 'SpaceNavigator' available for around $120

Would't using them require application support, like how DAWs support the Mackie Control protocol, among others?

If they are USB then they support HID

Not certain their use case of brush size has any relevance. Most graphics designers have a tablet and can easily set brush size to pressure control, or they just hold down the hot key that lets you adjust brush size by moving your mouse/pen so they already have analog input to brush size any time they want. Maybe some of the other professions they mention are more lacking in equipment, however. It is kind of weird talking about how hard it is to get to the brush size dialog, though, when no one uses that unless they are an amateur...

This argument sounds a lot like blackberry vs iPhone to me tbh. I frequently use tablets and yet setting brush size was never great, hot keys being the blackberry. I am very excited about this device, which feels like the right next step, the iPhone.

I agree. Although I believe it might have many uses in other areas, in Photoshop I am fastest with one hand on drawing tablet and other on keyboard. Having 3 modifier + all one key shortcuts/tools under my fingers is the fastest way. Specially in digital painting I might sample color, change size, paint, pan, erase, sample, reverse color, zoom; new layer all within 10-20 seconds. (alt; alt-right click pen; b; left click on pen; e; alt; x; alt+left click pen; ctrl+shift+n)

brush size? even faster than flow is alt+pen right click and hover pen on tablet.

Edit: I would thin it would be really neat music controller. I could see use in there(specially in party's)

The first thing I noticed from the video was how delicately people were using the device.

Not sure a majority of the developers I know would be able to be as dainty with something like this. Most type with heavy hands, beat up their mouse on a regular basis and have many keyboard shortcuts to reduce dev time in their Adobe programs.

For these people, using something which looks and is demonstrated as being very delicate, would not last long in their hands.

Having said that, how durable is this?

one of the founders here. Thanks for the comment. It's pretty durable. It's made from turned metal and acrylic glass. It will survive drops from tables and walls. It is mainly aimed at people who need sensitivity and pixel-perfect controls in graphic design, 3D modeling or video editing. But it is built on an open platform so it can also be used for things like music.

Looks like a more flexible (and wireless) version of Griffin's PowerMate[1]. The last hardware I crowdfunded is still mired in delays, but I like this idea enough to try it again.

A couple of questions, if you don't mind:

* Roughly how heavy is it? For physical objects like knobs, I like a bit of heft. Not as big a deal here as in a SpaceNavigator where it's pushing back instead of spinning freely, but I'm curious.

* Specs mention a 4 month battery life. Is that a AAA where I can use a rechargeable, or some sort of button cell? Or not yet finalized?

* How does software support work? I see it has smooth brush size control in Photoshop, but can I expect it to work that well in less common software like Leonardo or Mischief? Or will it have to be bound to keyboard shortcuts that change things in large steps like the square brackets in Photoshop?


[1] http://store.griffintechnology.com/powermate

thanks. Would love you as a backer. Please let me know which tools you want this to work with and we'll implement it. powermate: Yes, we have used it. We like it. It was too limiting for what we needed and we wanted it to be more portable and other materials. Weight: roughly 300 grams. That's one of the reasons we used metal, to make it heavier. We're also using a slip-resistant material on the bottom so it won't slip around. * Battery: We have worked built prototypes with both batteries, AAA and coin cells. This is one thing we're focusing on right now. We will definitely post an update on the campaign asap. Your thoughts from the user side would also be appreciated. * Software: The first and easiest form is using keyboard shortcuts and mapping it to an application. The second form is through scripts like Apple scripts.

I have a BC-700 charger and a bunch of low self discharge AAs and AAAs that I use in practically everything (game controllers, wireless mouse, TV/roku remotes, headlamp, and things I'm probably forgetting), so put me in as a vote for AAA.

For the slip resistant material, I don't know if it exists, but something that isn't a dust magnet would be great. I've noticed with my SpaceNavigator that the ring of rubber on the bottom tends to accumulate dust and start slipping until I wipe it down periodically. Again a bigger deal for them, since their 6-axis devices include pushing sideways as an input.

As far as software support, Mischief does 1px increments with [], so that one's easy. And open source programs like GIMP and Blender could always be modified if needed. So I guess I'm not too worried about software.

This is a kind of niche thought, but I wonder if I could suction cup it to the back of a Surface Pro. I tend to sketch with the tablet in my lap and no good place to set a Flow. Chairs with no tables, flopped on couches, etc.

Went ahead and backed it! The $20 worldwide shipping is a bummer, but I'm glad it's been priced into it. IIRC that's one of the big problems that Apollo Pens ran into.

Thanks you for the thorough response. Really valuable and thanks for backing us. I would love to stay in contact and talk more in detail about the suction cup, software etc. Please feel free to send me an email: hi@senic.com and we can take it from there.

  Looks like a more flexible (and wireless) version of Griffin's PowerMate[1].
Well, more flexible at least. http://griffintechnology.com/powermate-bluetooth

I used to use the wired version of this one. More specialized for 3D navigation, but pretty configurable, just like the rest. http://www.3dconnexion.com/products/spacemousewireless.html

I also use the wired SpaceNavigator. For a long time they weren't usable at all outside 3D apps. Prior to v10 of the drivers, they were adamant that all software developers should support it natively instead of letting it emulate joysticks and other already supported devices. Definitely got more useful when they added that.

It's a fairly different piece of equipment though, I'd compare the SpaceNavigator to a joystick or thumbstick (with 4 extra degrees of freedom) moreso than a knob, despite looking like one. I've never had much luck getting a balance of speed and accuracy with it, movement either feels slow, or you crank the sensitivity up and it becomes hard to do small motions accurately. Kind of like a gamepad as compared to a mouse; you have a quarter inch of motion and a few degrees of tilting to work with, and you can't map all of the different speeds you'd like to move at into that space.

I like knobs because they allow for a great mix of speed and precision. I know what half a turn does on my computer's volume knob, and I can do it a lot faster than "push the spacenavigator sideways a bit, hold it there until the volume is where I want it, then let go."

Which isn't to say the SpaceNavigator is a bad device. It just serves a different purpose.

IMO Flow's biggest competitor is the touch wheel on Wacom's Intuos line. Among professionals who spend time in Photoshop, they're extremely common. No such wheel on a Surface though, and having a few shortcuts that I can map without hooking up the keyboard would be a nice addition. Hoping this pans out!

sorry for the messed up formatting.

Question -- why does your campaign involve "EMV" & FCC certification? Do you mean CE?!

Yes, CE certification is needed as well. EMV is a German thing - electromagnetic compatibility tests which are needed for consumer electronics.

And about FCC. This article explains it pretty well: https://www.sparkfun.com/tutorials/398

I am fully aware of CE and FCC. As a hardware engineer I've had to obtain those for a number of products myself.

Could you please point me to EMV for Germany? I've never heard of it before and am fairly certain CE covers Germany as well as all of Europe.

Do you mean TUV by any chance?

The TÜV can do these EMV tests for you. But it's a standard in whole Europe. It's part of the CE marking. Not too easy to find literature on this in English. I hope this helps you: http://ec.europa.eu/enterprise/sectors/electrical/files/emc_...

Haha. Those are EMC tests, not EMV.

I've never heard anyone refer to it as EMV. Is this perhaps a German language thing? Does Electro Magnetic Compatibility translate to EMV in German?!

EMV is Euro-Mastercard-Visa. :)

Hehe. Yeah. As said, it's a German thing. EMV stands for "elektromagnetische Verträglichkeit" which translates to EMC.

Haha. Gotcha! Learned something new. :)

When you said "a German thing" earlier I thought you meant there is a specific standard for Germany called EMV. I was so confused.

Love the idea! I'm always looking to see if any gesture recognition controls will take off for computers, this seems like a very good one as it doesn't involve raising your hands to any excessive degree. There have been gesture devices in the past, but they all seem to want you to raise your arms and wave them about in large-ish motions. Good to see a device that I can still rest my elbows on the desk and just wave at! I can see this becoming like a 2nd mouse to the left of my keyboard, a power device for certain apps.

I will just add one slight tip if I may. The first two sentences of this page don't really explain what the product is:

"We (YC S13) just launched our product “Flow” on Indiegogo. It’s a programmable and intuitive wireless controller that that gives you high precision and speed."

I was left wondering what it was a controller of? An RC plane? A games console? Industrial machinery? Just add "controller for your computer" and it'll read a lot better. It's not helped by the fact that the gif below only shows a hand waving over a circular thing with no visible feedback from a computer screen ;)

Done. Thank you for your feedback!

Aw, an advertisement blog post disguised as special insight.

Plenty of comments here on how this problem has already been solved by other devices.

This one appears to be different by allowing functionality to be extended more easily via programming. 3dconexion spacemouse for example is not that easy to interface although 3rd party programs like Glovepie do allow interfacing via simple scripts.

Wacom tablets would seem to be the most obvious existing product out there in a similar space - simple scripting can interface Wacoms with other apps like music players, so perhaps this new device isn't necessary.

Users in the video are interacting delicately, this looks like a problem - when you are zipping through interactions to achieve tasks (especially when they are repetitive) you want to be able to have good repeatability. Delicate movements here are your enemy.

I guess the TLDR of my comment is that nothing out there really solves all the issues out of the box and this one might not either unless it can build an ecosystem of drag and drop scripts for people to use for their day to day tasks.

edit: interestingly in the list of "requirements" (para 2.) the ease of installing new abilities or scripts or whatever extensions is not listed other than "it has to work with the tools we use" and in the "developers" section it should mention an "app store" otherwise no-one's gonna find that awesome Photoshop extension for it.

This looks worse than a mouse. I'm just trying to imagine how I'd use this in a game like Call of Duty or Team Fortress. The modern optical mouse gives you precise movement and acceleration in 2 dimensions. The touch surface of this thing is too small to work better than a mouse in a game situation unless you crank the sensitivity way up at the cost of accuracy.

At first I thought it would be movable and contain an optical mouse tracker underneath it, but was disappointed to see it's basically a trackpad with a wheel and a gesture sensor.

Maybe it'll work good for many apps, but it seems every time I see something that tries to improve on the old mouse, it ends up looking worse, at least for my use case.

I'm not sure that this is meant to be a replacement for a mouse -- seems more like an additional tool for specific tasks (like the Photoshop task mentioned in the article).

I really want more things like this to exist, but I can't say exactly why or what I would use them for.

I hope this thing will succeed, but I can't imagine what the exact market is or whether its big enough to sustain this project.

That is one of the major changes that are happening. * Existing user interfaces (screens, mouse, keyboard) have made sense because they were the lowest common denominator and because prices could be kept low this way. * Now, prices for hardware are dropping (we will soon have the first wifi chip for less than 1 USD.). That means that we can explore completely new user interfaces, ones that are perfectly designed for the person and their needs. You don't need to have a user interface anymore that is designed for billions of people. You can manufacture several kinds of user interfaces quickly and that suit millions of people, just better. * Manufacturing is also becoming much easier and cheaper. There are more and more robots that can do complex work. They are also becoming cheaper and more capable which also influences this aspect. * I believe that the future of user interfaces does not lie in centralized, generic devices but distributed, specific interfaces.

I really want more things like this to exist, but I can't say exactly why or what I would use them for.

I feel the same way and have enough spare cash to back it. Twald's explanation makes some sense, but really I think this is the essential part: "That means that we can explore completely new user interfaces, ones that are perfectly designed for the person and their needs," because I translate it to mean, "We don't entirely know what'll end up happening."

I'm happy existing in that state of ambiguity, however, and for $100 it seems worthwhile to find out what the future might hold.

We have a few products in mind that would instantly add value. Think about your feet or objects and surfaces around you like a table. Why not use them? Why not blend technology into natural things that use more senses than just your eyes and a reduced form of haptics.

Reminds me of the Monome Arc which apparently isn't being built anymore, but here is a story about it http://www.engadget.com/2011/01/24/monome-arc-osc-controller... ... of course this adds the touch pointing and non-touch gestures which seems pretty cool, if they can deliver the thing.

Article aside, I don't really understand how this is a cool (or even necessary) device... I watched the videos on their Kickstarter page, and correct me if I'm wrong, but it has three functions that I don't need: 1) The outside rotation for adjustments; Mouse wheel? 2) Basic swipe gesture recognition; Um, any button, and I don't have to be hands off the keyboard? 3) Touch tracking; Trackpad...?

I'll definitely buy this when it ships in 1-2 years.

One of my coworkers has a Griffin Powermate[1]. It's corded and the software is a bit clunky but is quite similar, although it functions a bit more like a mouse. [1] http://store.griffintechnology.com/powermate

I was about to comment about that... I've had one for years and still find clever things to do with it. For example, when searching a project in my editor for a string, turning the knob jumps between search results.

It's amazing how much functionality you can eke out of such a simple input device.

It's made me consider other purpose-built init devices (like some of the midi controllers mentioned earlier). I have a Native Instruments controller designed for one of their massive sample libraries that I'd love to hack as a general purpose programmer's helper.

Anyone have any experience with such devices? What other clever uses of these alternate input devices am I missing out on?

Smartphones are largely unexplored as a secondary input devices. It can be a input for gestures (zooming, dragging, etc), as an extra place to add custom keys (think macros), custom sliders and other touch-based UI controls, or it can be a notification device ("compilation complete")

This could also be a nice alternative to a steering wheel for driving/racing video games.

They totally avoid saying it's an optical mouse:

"changes in position are measured by sensing the scattered laser light which is reflected by an object."

cool device, i will buy it.

Because I skim through a lot of articles at first to determine if something is even worth reading. I thought this was how the task management app Flow [1] was built. I thought they where trolling with some of the hardware and "Angular" stuff, then I realized sometimes skimming isn't the best approach.

[1] https://www.getflow.com/

...while I thought it would be covering Facebook's Flow project posted here a week ago: https://news.ycombinator.com/item?id=8625222

Could also be confused with Square's Flow library.


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