Have Other User Intefaces Been Considered ?

Paul, WB5AGF

Over the past 15 (or thereabouts) years there have been several radios (both receivers and ham transceivers) which have used (or at least allowed) computer based (PC) user interfaces (UI).

This started with analog radios which, although they had a 'real buttons-and-knobs front panel', on the backside they had an interface that you could plug into with a computer if you wanted to. Then we had some analog radios that didn't have the buttons-and-knobs front panel (the Kachina and Ten-Tec Pegasus come-to-mind) but instead required that you interface with them through a PC UI. Then we had radios which began to use digital signal processing (DSP) and some of those off-loaded at least some of the processing to the PC and had a 'thick' inteface UI (the FLEX-5000 for one). Now we have radios which are (almost) entirely 'digital' - having co-located dedicated processing hardware (e.g. the FLEX-6000s) with a thin-EtherNet client UI used to take the place of the buttons-and-knobs that many of us grew up with (eons ago ....).

For the most part, ever since we began to move away from the physical buttons-and-knobs front panel, the PC UIs have taken their cue, on how they should look, from the previous physical front panel interfacing radios .... until the FLEX-6000 radios.

I confess .... I'm having trouble with the buttons-and-sliders on the right side of my SmartSDR UI screen .... with most of the real-estate being taken over by the panadaptor/spectrum analyzer  display (I like the word Panadaptor ..... reminds me of when I was a kid - given a 1950s era Radio Products Panadaptor - got it working with my Lafayette HE-10 shortwave receiver - learning to adjust the IF amp's frequency response gave me an appreciation of critical-coupling and bandpass tuning).

I do not want to start a fuss. I'd just like to know if there are others who look at the Perseus UI (and similar) and think 'Ah ....'  then the thought .... How hard would it be to use the API to build something similar for the FLEX-6000 ?

(Now I'll be quiet .... don't want to cause-a-stir .... just wanted to ask.)

- Paul, WB5AGF

Jim Jerzycke

Interesting idea. A new hardware interface to the 6000-series that gives them the look and feel of a "knob-and-button" radio.

73, Jim

Bill W2PKY

Check out Android App on this forum.

Jim Jerzycke

I don't think Paul means another "app" with a GUI. He wants a new box that plugs into the Flex, and gives you *real* knobs and buttons rather than another virtual front panel.

73, Jim

Jim K4JAF

Are referring to another type computer screen interface or a "knobs and buttons" physical interface?

In either case, don't think there would be that much demand to justify the expense.

Jim K4JAF

N7AIG

I remember the initial shock and a profound sense of loss, way back, when our new PDP-11/32 arrived with only a key on/off switch on the front panel.... what happened to all the toggle switches and lamps !?

Robert -- N5IKD

Many people are experimenting with using MIDI controllers with micro-controllers for this very purpose. https://community.flexradio.com/flexradio/topics/flex-6000-with-midi-controller Others are using iPad and Android apps for this.

Walt - KZ1F

Well, there is the work William is doing. What I am doing is a new and replacement for the FRS FlexLib. The UI will be runnable everywhere and take the best parts, well, what I perceive as the best parts of the SSDR plus some missing features. After that process is complete I will create an HTTP based UI That will be markedly different. While some like the panadapter concept it is too married to the single application model. I will be removing that restriction and, indeed, making both the spectrum display and waterfall option. There is no reason why a smartphone display would not suffice to adequately handle a qso, spot to log entry.

Paul, WB5AGF

(This is a good group of responses ..... I'm going to post responses to each.)

Jim;

In my mind's-eye I was 'seeing' the Perseus's UI - I've recently been looking at how several radio UIs have been implemented.

- Paul, WB5AGF

Steve W6SDM

I am using the Behringer interface.  It's not that I miss buttons and knobs - I don't.  It's just a fact of Windows life that you need focus.  When operating a contest, or split DX pileups, having instant access to functions is a real advantage.

Paul, WB5AGF

Bill;

I you saw my cellphone you'd start laughing ..... 'being' a 3G phone is all that it can manage.

A good friend (ham of course) and his wife are both iPhone owners .... when I'm with them somewhere I find out which one has upgraded most recently (I should buy some Apple Stock .... its gotta' go up) .... We can be having lunch on a Saturday, or going to a hamfest, and if I innocently make the comment "I wonder ... ?" out comes the iPhone and it's 'Google-to-the-rescue'.

- Paul, WB5AGF

Paul, WB5AGF

Jim;

Actually when I posted my original inquiry I was thinking of a software UI that would be more in-line with other radio's UIs (e.g. the Perseus).

However .... it's interesting that you made the comment you did ..... I've actually wondered (day-dreaming ?) at the practicability of putting together a physical UI (a 'box') that would (as you described) plug into the FLEX-6000.

(Back in the late 1980s I bought a Ten-Tec Paragon Transceiver and I've always thought that it has as 'sanitary' a radio control panel as I've ever seen. And then of course there' s the National HRO-500 Receiver .... goodness they're pretty.... <g>)

Just imagine a 'box' that looked like the Elecraft KX3 plugged into a FLEX-6000.
(One of the guys that I see every September at a hamfest in western Arkansas has a KX3 - the first time I saw his - it was all I could do not to call Elecraft and buy one.)

- Paul, WB5AGF

Paul, WB5AGF

Jim;

I was a RF/Systems Engineer (back-in-the-day) and don't have the experience to know for sure what's practical to implement in a software-based UI.

My question (roughly speaking) was wondering-aloud how difficult it would be to use the FLEX-6000 API to put together a UI that is a bit less 'innovative' (i.e. more 'traditional') than what's currently provided ?

If I don't have any serious misunderstandings about what's implemented in the FLEX-6000 hardware then I suspect (but don't know) that there's enough horsepower-under-the-hood such that a minimalist interface can be implemented without-reinventing-the-wheel.

- Paul, WB5AGF

KY6LA_Howard

Paul

There is already an Open API which provides the data from the radio in VITA49 format.  It's up to you to decide how you want to display it... I might add that they have not currently opened up the internal Display API but I am betting that will happen too...

Paul, WB5AGF

A friend (ham) in Wisconsin had an Altair 8800 back in 1980 .... I had trouble understanding that anything useful could be accomplished flipping those toggle switches back-and-forth.

I remember when we got VT-220 terminals on-our-desks (now that was exciting).

- Paul, WB5AGF

KY6LA_Howard

I go way back before that ... We programed our computers with patch panels.. It was a miracle when we got a teletype and paper tape to load machine language code.  Our computers were so slow, that if a tube burned out, I could grab it with a glove and change it out before the program lost its next step

KY6LA_Howard

I use a KX3 sometimes at the NX6T contest station. Great receiver.

But like the K3 you really need to be EXTRA careful not to push the wrong button or hold a button too long or double click a button.. it then becomes a UI Nightmare with multiple nested menus to page thru.

Frankly I like the very clean SSDR visual interface

That said I have seen several interesting graphical UI.. for example

1. NaP3 has a really functional panadapter that shows spots and worked spots on the waterfall....  something we desperately need in SSDR.

2.  the FMD-DUO UI uses different colors for different slices so you can quickly distinguish them..

there are many more

Paul, WB5AGF

Robert;

I've looked at K6TU's Apple app to see if it's clear to me (I want to understand) that all critical whistles-and-bells are there.

Makes me wonder what it would take to provide a similar control interface as a PC UI ? (Or, since it runs on portable Apple devices .... How hard would it be to port the app to OSX ?)

- Paul, WB5AGF

Paul, WB5AGF

Walt;

Is it feasible to implement a FLEX UI under LINUX, and by virtue of that have it functional on both native LINUX boxes as well as under OSX (a friend, another ham, tells me that he has had reasonable success at porting LINUX applications to run under OSX) ?

I believe that it is correct to say that anytime a software application is run under an OS other than what it's native to .... that processing speed suffers .... but I've been reading that there are people who believe that PC hardware is now fast enough to allow running virtual OSs for everything (multiple OS 'windows' all open on the same PC desktop ?).

- Paul, WB5AGF

Paul, WB5AGF

Now Howard come on .... Next thing you'll be telling us is that you were a midshipman at Bletchley Park .....

KY6LA_Howard

Not that old but it would have been fun. Built my first computer in April 1958. 4 bits 8 tubes. Been playing with them ever since.

Paul, WB5AGF

Howard;

Ya' got me beat ....

I believe that was about when I taught myself that there was such a thing as series and parallel circuits while figuring out how to set off firecrackers electrically using the stranded steel wire in the battery powered phones that my folks bought me. (I ran the phone line from the wall in my mom's kitchen outside and up into the trees that were my imaginary 'ship' .... I 'sailed' the South Pacific from the pine trees in our backyard .... the TV show "Adventures In Paradise" was my favorite.)

- Paul, WB5AGF

KY6LA_Howard

I had a cousin who was a physicist who was working on the very early computers in the 1940-50's  He caused me to become really interested in computers. He would let me wire up program patch panels for him which I thought was really cool.  

So in 1958 thinking I was the high school computer maven, I built a 4 bit 8 tube computer for a Science Fair.  My physics teacher said "Blah"... ( i stole that design from Popular Electronics)

So I built 2 - 4 bit computers and hooked them together with wire to send counts from one to the other.... My physics teacher said "Blah"

So I built a single tube transmitter and used my Mom's 1935 AM Radio, connected them to the respective computers and sent data counts wirelessly over the air between them.

My physics teacher said "Amazing...But it's illegal"...

"Why?" I said, 
He said;  "You need a Ham License to use that transmitter"

"What's a Ham License?" I said

I got a Ham License, won the Science Fair which got me far too young admission into engineering college and the rest is history. 

Joe, KQ1Q

WB5AGF: "...ever since we began to move away from the physical buttons-and-knobs front panel, the PC UIs have taken their cue, on how they should look, from the previous physical front panel interfacing radios .... until the FLEX-6000 radios."

There is actually a furious debate within the UI design community about this exact topic. Namely whether real-world shapes, textures, indicators and controls should be slavishly replicated in the computer UI. Those are called skeuomorphs, and the new approach is called "non-skeuomorphic design".

The Perseus UI and other discussed knob radio computer UIs are skeuomorphic. SmartSDR is a step toward non-skeuomorphism. I don't know if this was a formal intent by the designers but that was the result.

The older skeuomorphic approach often uses screen real estate for essentially decorative elements. These are disparagingly referred to as "chrome" by the non-skeuomorphs.

There are stronger and weaker forms of non-skeuomorphic design. A weaker form might be "let's not blindly replicate real-world UI elements if we can think of a better way". A strong form might be: "every real-world shape and texture must go, absolutely no color gradients, no drop shadows, no chrome of any kind." It is sometimes called the "flat look" UI. Microsoft and Apple are both moving their UIs in that direction.

There have been design excesses on both sides. Apple's old Mac calender which had a stitched leather binding, etc: http://www.fastcodesign.com/1670760/will-apples-tacky-software-design-philosophy-cause-a-revolt

OTOH the non-skeumorphs produced the flat, tiled Windows 8 UI. In their zeal to jettison all chrome, gradients and drop shadows, they have sometimes produced UIs which are non-intuitive, hard to read, and difficult to use.

I personally like the SmartSDR interface, even though it's not perfect. It feels uncluttered, clean and fresh, and puts the big panafall display front and center. If only they'd implement hotkey commands for common controls, you could operate it with all the controls minimized and off screen.

Carl Moreschi

Take a look at my N4PY Pegasus Plus UI.  It's at www.n4py.com

DrTeeth

I also like the SSDR interface on the whole. It is, however, riddled with minor inconsistencies  and annoyances. Once 1.5 is out of the door, I would like to see the release after that to address the UI.

Walt - KZ1F

Paul, that is, what you just said, the necessity, being the mother of the invention. Yes, my primary goal is to have a smarter SSDR running on Linux. As it happens, it will also run on Mac and Windows and Raspberry Pi. In theory, it should work on Android but there are other things I plan for Android so it may get a unique UI and even more unique features.

Your second paragraph is kind of wrong, well, depending on what you really meant. If you are strictly talking about Parallels or KVM or VMWare, it depends. KVM, of which I know the most actually, depending on the app, could run faster in a virtual machine. I am not doing a virtual machine implementation, these will be native apps and, absolutely, run faster than a virtualized Windows environment running SDR. There is a real world case where rather than buy every employee their own machine, they get a virtual version of what that machine would be running. It's called virtual desktop Infrastructure. That provides way better security and cost structures to companies. You've likely heard a lot about 'the cloud'. That is, essentially, what in a business environment you are talking about. There is about a 10:1 ratio between a host machine and the number of virtual machines it can support. When you talk about something just a degree or two away from that, 'containers' there is a 100:1 ratio. However, what I am doing won't require virtualization. It is a single app that will natively run on Linux and many other platforms, even Windows. Having said all that, the one can I have, to date, kicked down the road is DAX / virtual audio channels. At this moment I do not know how to create a virtual device in Linux. It's not an impossible task, just not one I taken on as yet. The other issue is DAX is used to logically connect HRD or CW Skimmer to SSDR. Where they are all Windows applications connecting to them from not Windows is problematic. However, the better design is to connect to other apps through UDP/IP, which is how the IF data gets to SSDR to begin with. Why convert that to an OS audio device just to have it converted back to raw data. I believe FRS is resolving this problem.  What I am doing will directly handle spotting, logging, and LOTW uploading. I don't mean, necessarily, one gigundo application. As part of this effort I will spawn off service endpoints, singularly focused services with a discrete internet address and requests will be made to it. It will be blindingly fast.

As far as what your friend said, yes, one can port a C or C++ program from Linux to Mac. There are other languages where no port is required. In the case of S(er)SDR the only difference between the versions, Linux and Mac, will effectively be the GUI itself will look more like Mac apps and not Window or Linux apps. But beneath the appearance will be identical code. This will be very exciting.

Walt - KZ1F

Stu would obviously be the best person to ask but my understanding is Stu wrote a contest centric UI. Which is actually smart. If all you want to do is work a contest, you don't want all the fru fru stuff muddying up the water vs. laser focused on the one goal fast efficient contesting.

Paul, WB5AGF

Robert;

I didn't understand the sirens' call of MIDI Controllers until I began to read (and look) at the FLEX forum threads (what you referenced and : https://community.flexradio.com/flexradio/topics/hercules-midi-with-6300).

When I originally saw the term MIDI in the FLEX forums I remembered years ago when people were creating really clunky sounding imitations of piano music (or Stars-And-Stripes-Forever via MIDI; shudder) .... I was not 'current' and did not realize that the MIDI interface has spawned several physical interface 'boxes' .... now I see the attraction.

(True Confession - I still think it'd be neat to put a 'thin-box' together that looked something like the Ten-Tec Paragon's front panel but with a circuit board behind it that interfaced with the physical front-panel controls and then provided the data interface back to the FLEX Radio.)

- Paul, WB5AGF

Carl Moreschi

That's roughly the idea behind using the Tmate2 with my N4PY software.

Walt - KZ1F

In that regard, that would be an outstanding move for FRS to undertake. Move away from the single monolithic **** client app. In other words, 1) have each pan be a child of the desktop not a child of the HWND_MAIN (app). Right now the main app is, for lack of a better word, using a border layout, there is are the buttons, BAND, DAX, etc on the left, the VBox on the right, menu across the top and whatever on the bottom. What takes up the most real estate is the pan in the center. If a user could move that off to another window the main app real estate reduces dramatically. 2) In the tabbed VBox on the right, using a small form like a laptop not all the dialog boxes are visible, in order to get to EQ I have to close several others. They should be scrollable. Or, ideally, let the user put the pan and the TX dialog in front of him and the others off line of sight.