Third time lucky

  • 3
  • Question
  • Updated 3 years ago
  • (Edited)
After the release of a second version of SSDR that does not have much for non-Maestro users, I can only hope that the next one can redress the balance. I think the Maestro users have had their fair share of expensive engineering time for the moment.
Photo of DrTeeth

DrTeeth

  • 1687 Posts
  • 389 Reply Likes

Posted 3 years ago

  • 3
Photo of Jd Dupuy

Jd Dupuy

  • 155 Posts
  • 60 Reply Likes
Funniest thing I have read all day! Now we have 2 camps, Maestro users and non-Maestro users.
Photo of Jay / NO5J

Jay / NO5J

  • 1674 Posts
  • 262 Reply Likes
At least 3 camps then, You overlooked "The my way, or the highway dammit!" camp.

73, Jay - NO5J
Photo of John-K3MA

John-K3MA

  • 102 Posts
  • 29 Reply Likes
Should I assume your not buying into the explanation that "SmartSDR improvements made for Maestro and the Contest users benefit all of us"?
Photo of KY6LA - Howard

KY6LA - Howard, Elmer

  • 3762 Posts
  • 1621 Reply Likes

@GUY

Sorry to BUST YOUR BUBBLE

But the following were NOT MAESTRO related SSDR Improvements in 1.8.3 which means that the Second Time is ALREADY the Charm.


The cumulative SmartSDR v1.8.3 ecosystem Change Log detail is listed below.

-------------------------------------------------------------------------------
 v1.8.3 2016-06-30
-------------------------------------------------------------------------------
General release of SmartSDR for Windows.

-------------------------------------------------------------------------------
 v1.8.2 RC 2016-06-24
-------------------------------------------------------------------------------
#3622    Firmware: Mic audio will transmit even if DAX is enabled unless DAX is actively streaming.
#3718    SmartSDR for Windows: Installer no longer installs DAX drivers when the DAX option is unchecked.
#3684    API: Fixed Vita49 info class typo.

-------------------------------------------------------------------------------
 v1.8.1 Alpha 2016-06-16
-------------------------------------------------------------------------------
#3597    Firmware: Various fixes under the hood to address random radio crashes (sometimes with ~400 Hz tone afterwards)
#3666    SSDR for Windows: changed radio connection lost message if the client was disconnected by another user.
#3673    SSDR for Windows: radio upgrade failure message is now more descriptive.

-------------------------------------------------------------------------------
 v1.8.0 Alpha 2016-05-27
-------------------------------------------------------------------------------
#3505    Firmware: Users can now disconnect users that are currently connected to a radio.
#3644    Firmware: Added automatic power foldback for high SWR situations (starting at 25 W reverse power)

-------------------------------------------------------------------------------
 v1.7.34 Alpha 2016-05-25
-------------------------------------------------------------------------------
#3657    Firmware: Fixed missing CW sidetone.
#3610    Firmware: Waterfall black level is now restored properly on startup.
#3566    CAT: Added checks to prevent UHE when disconnecting.


Photo of Walt - KZ1F

Walt - KZ1F

  • 3040 Posts
  • 645 Reply Likes
How many of those fixes fixed defects introduced by 1.7?
Fixing defects is good.
Not introducing defects, priceless.
(Edited)
Photo of EA4GLI - 8P9EH - Salvador

EA4GLI - 8P9EH - Salvador

  • 1782 Posts
  • 547 Reply Likes
There is also the deep-pockets vs completely-broke-no-clue-why-they-bought-a-flex-radio camps.
Photo of SDR Fan

SDR Fan

  • 10 Posts
  • 2 Reply Likes
Gee I hope when you pay your subscription your $200 doesnt brick your radio
Photo of EA4GLI - 8P9EH - Salvador

EA4GLI - 8P9EH - Salvador

  • 1782 Posts
  • 547 Reply Likes
My rhetoric is hitting nerves.... :)
(Edited)
Photo of Jay -- N0FB

Jay -- N0FB, Elmer

  • 539 Posts
  • 213 Reply Likes
"The camp town women sing this song...Doo Dah, Doo Dah!"
Photo of Jay / NO5J

Jay / NO5J

  • 1674 Posts
  • 262 Reply Likes
Camp Reality may not be for everyone, But I find it to be a little more efficient

YMMV!

73, Jay - NO5J.
Photo of DrTeeth

DrTeeth

  • 1687 Posts
  • 389 Reply Likes
Howard, those non-Maestro fixes are quite esoteric and more importantly, it is the ratio of Maestro to non-Maestro fixes that stands out a mile. I was told by a reliable source a while ago, March to be exact, that there were over 3000 bugs on the Flex bug tracker, many of which had been squashed, but "many, many more to go". I do not recall how many releases we have had since then, but I do not recall hundreds of non-Maestro bugs being fixed since then. I take "many, many more" to be well over 1,500. I will not divulge my source under any circumstances, but would be tempted if Flex were to offer to buy back my Flex at a decent price so I could start again.
(Edited)
Photo of Gerald - K5SDR

Gerald - K5SDR, Employee

  • 830 Posts
  • 1514 Reply Likes
Guy, your "reliable source" is totally unreliable and the numbers quoted are categorically false

Also, our software tracking system tracks everything including virtually every feature request ever made along with many things we think of that we might like to do sometime in the future.  It even includes large numbers of duplicate entries reported by different customers, cockpit error problems, and bugs that have already fixed.  We have to do periodic scrubs to clean the backlog list from duplicates and bugs that are already fixed.  We are due for a full scrub soon but that takes significant time away from programming.

We use Agile Development methods to set priorities on a real time basis based on the most important overall priorities within the release cycle.  I make no apologies for doing one and a ~half software releases on Maestro.     
Photo of Ernest - W4EG

Ernest - W4EG

  • 215 Posts
  • 39 Reply Likes
Thanks for the info Gerald on how you keep track of reported issues.  73
Photo of DrTeeth

DrTeeth

  • 1687 Posts
  • 389 Reply Likes
Gerald, if you knew who the source was you would be as surprised as I that the information was way off. Thank you for your correction.
Photo of Gerald - K5SDR

Gerald - K5SDR, Employee

  • 830 Posts
  • 1514 Reply Likes
No, we don't do MVP per Agile.  That, I believe, is more applicable to web based products and smart phone apps I believe.  However, in software there is always more you could add to a release so you have to cut it off somewhere.  Otherwise you would never ship anything. Everything is a trade off.
Photo of Walt - KZ1F

Walt - KZ1F

  • 3040 Posts
  • 645 Reply Likes
It does in the way Agile has been implemented at several large corporations (CA Technologies, Monster.com) I've been in the development unit of. However, the creation of MVP represents the confluence / interlock of product marketing, sales, and development to determine that set of features required for sales to hit their defined quotas that make the development and management costs justified to undertake the release. I don't know how that equates in a small company where those functions are less well defined. My experience is no two companies implement Agile quite the same, as it doesn't necessarily fit the culture or the management that is the primary recipient of loss of status. Obviously, I can not speak to how FRS has implemented it, but given the sizing differences it would have to be a hybrid, I would guess, Gerald, Steve, and Eric discuss it over a few scotches?

Yet it does work well once egos are removed. It works well in non "web or smart phone" environments as well.
(Edited)
Photo of Gerald - K5SDR

Gerald - K5SDR, Employee

  • 830 Posts
  • 1514 Reply Likes
Guy, while I won't give numbers here publicly your source is simply wrong. There is nothing even close to that in Github even if you include all duplicate entries and already fixed bugs.  

By the way, an "issue" in Github could be anything from a one line code fix to a scientific research project that could take months - and everything in between.  Some of them we may never do but we still track them as "someday/maybe."
Photo of Gerald - K5SDR

Gerald - K5SDR, Employee

  • 830 Posts
  • 1514 Reply Likes
Walt, your statement below is actually fairly close in principal to the way we do it.  Our implementation is ever evolving as we grow.
However, the creation of MVP represents the confluence / interlock of product marketing, sales, and development to determine that set of features required for sales to hit their defined quotas that make the development and management costs justified to undertake the release. 
Photo of WW1SS - Steve

WW1SS - Steve

  • 793 Posts
  • 268 Reply Likes
All I am going to say is I spent over 5 grand for this 6500, maestro, flex knob. There is no excuse for this and it should not happen to one single unit.. It is getting to the point that I am afraid to upgrade. If it happened to my rig I'de be pissed.
Photo of Phil m0vse

Phil m0vse

  • 208 Posts
  • 38 Reply Likes
Unfortunately this happens with software occasionally. If all of the Alpha/Beta testers didn't experience the particular set of circumstances that caused the recent Maestro bug it can slip though the net. I upgraded my Maestro to 1.8 before the issue was discovered and it worked fine (more by luck than judgement it appears). This is why many people wait a few days before installing any updates. I am just too impatient :)  
 
Photo of Walt - KZ1F

Walt - KZ1F

  • 3040 Posts
  • 645 Reply Likes
Gerald, I was thinking you might say, yep, there is this quiet little bar in the next complex that we hold our executive meetings at and scotch has been known to be in attendance on occasion. Certainly in the case of CAT, they have a massive development organization representing the bulk of the 16,000 (I think it's down to 12,000 now) employees. So the commitments made to upper mgmt before funding is released is taken quite seriously, so there has to be complete signed off agreement within that triad (marketing, sales, and development). Monster is far smaller than it was before I left..which is why most of us let it.
(Edited)
Photo of Gerald - K5SDR

Gerald - K5SDR, Employee

  • 830 Posts
  • 1514 Reply Likes
Steven, Phil is correct.  The Maestro update issue on 1.8 did not show up in any of Alpha/Beta test locations.  Same thing happened with my iPad a few releases of iOS back.  Apple did not catch a major upgrade bug until they had millions upgrading.  Only happened on some units.  Had to wait a couple weeks without use of my iPad as I recall till Apple released a fix.
Photo of Phil m0vse

Phil m0vse

  • 208 Posts
  • 38 Reply Likes
I feel for you Gerald, at least it doesn't affect every Maestro or you would have a major headache on your hands. I remember when Sophos AntiVirus released a patch which identified it's own updater as a virus and this was pushed to every user, requiring manual intervention. You couldn't get through to them on the phone for about 2 weeks!
Photo of EA4GLI - 8P9EH - Salvador

EA4GLI - 8P9EH - Salvador

  • 1782 Posts
  • 547 Reply Likes
In my case I was praising the update because it solved network issues I was having with the Maestro....
Photo of Tim - G7GFW / F4VQP

Tim - G7GFW / F4VQP

  • 170 Posts
  • 73 Reply Likes
I am sure that someone will give you a reasonable price for your Flex if you advertise it on HRD then you can buy something else to whinge about.
Photo of Jay -- N0FB

Jay -- N0FB, Elmer

  • 539 Posts
  • 213 Reply Likes
Steven, my Maestro was affected by the issue. While I'm not happy that it occurred, I certainly understand that in a software world, perfection rarely happens. General Dynamics, Northrup, Cisco, etc. all make products that cost multiple millions of dollars and they have all had catastrophic failures occur. When this occurs, it is how the vendor responds that makes the difference.

FlexRadio response to this issue is nothing less than exemplary. So Steven, if you want to be "Pissed" because you are expecting perfection, then your life is going to be full of frustration and disappointment. In the end, It's up to you as to how you want to react to your self defined expectations.
Photo of Steve - N5AC

Steve - N5AC, VP Engineering / CTO

  • 1057 Posts
  • 1097 Reply Likes
While I don't remember reading about the concept of MVP, I do think this is exactly what we do with the Alpha team.  We discuss regularly "do we have enough for a release."  To us, this equates to the question: "Is there enough new software to justify us cutting a release and the Alpha team having to install and test new software."  This is not, however, what gets released to the public.  

Immediately after a public release, we fit in all kinds of high-risk things we've been working on in the background.  We just went through a lot of this today and one of our developers said "the is the highest risk release we've had in a long time."  But this is our model -- the later we get in the process, the less risk we add to the software.  Then when we release to the public all of the high risk stuff has been baking for a long time and the low risk stuff may be recent.  At the start of the next release, we add in all the high risk stuff again.
Photo of Gerald - K5SDR

Gerald - K5SDR, Employee

  • 830 Posts
  • 1514 Reply Likes
Guy, I think I just realized (with help from Steve and Tim) how someone could make the mistake of thinking there are over 3,000 outstanding issues.  We number everything coming into our tracking system with a sequential number from the "beginning of time."  I think we are at somewhere around 3,700 sequential issues since t=0.  

The number includes all development components for SmartSDR since the beginning, the entire FLEX-6000 Series development, Maestro development, CAT/DAX development, support for third party developer enhancements, some PowerSDR issues, commercial product development, etc., etc.  There are even hardware issues, testing, and documentation tasks in that sequential number. The number has no direct correlation to open issues because it includes everything that has been closed/completed in the past.  Individual issues might have taken months to develop for new features and some might have taken minutes.  

Context is everything.  Saying 3,000 issues is like saying the temperature is 0 degrees.  It all depends on the system of measurement.  Is it degrees C, F, or K?
Photo of Walt - KZ1F

Walt - KZ1F

  • 3040 Posts
  • 645 Reply Likes
This is weird as something like this just happened to me except your dbl post looks like a true dbl post (note the two 'likes'. What happened to me the other day was 1) what looked like a dbl post except the first one didn't 'terminate' with a like etc, it just had the separator and there was the second one, and, 2) it absolutely errored out when I tried to delete (remove) the second one. The whole 'red bar' thing. I even got out and came back in, in case it was state information cached in the browser. This is new.
Photo of Walt - KZ1F

Walt - KZ1F

  • 3040 Posts
  • 645 Reply Likes
3719, but, no, I am not counting, Howard volunteered that number.
(Edited)
Photo of Walt - KZ1F

Walt - KZ1F

  • 3040 Posts
  • 645 Reply Likes
Yet it does work well once egos are removed. It works well in non "web or smart phone" environments as well.

Kevin, the thing to understand is MVP is MINIMAL viable product. What development agrees to do in any given sprint is up to them. That's why I mentioned egos as the scrum team determines what they will sign on for, not the development manager, so it is a loss of status for them and that causes, understandable, resistance to their career changes. The team could agree to do more but, in my experience, they generally agree to less and the MVP line is redrawn. Kind of like kids with the "I dare you to cross that line" only to end up redrawing the line and repeating the challenge. In the case of CI where releases flow every couple of weeks and a sprint is maybe 2 weeks long followed by a targeted release, that does apply more to the Monster.com / smartphone environment. CA Technologies has a huge global development organization and massive software portfolio, precious little of it, percentage wise, is in the smart phone and web based world.
(Edited)
Photo of Steve - N5AC

Steve - N5AC, VP Engineering / CTO

  • 1057 Posts
  • 1097 Reply Likes
For the record, this doesn't describe our process at all.
Photo of Walt - KZ1F

Walt - KZ1F

  • 3040 Posts
  • 645 Reply Likes
No, I don't imagine it does.

This conversation is no longer open for comments or replies.