I had resigned myself to the idea that the QSX project from Hans G0UPL had proved overwhelming - thus the long delay and perhaps cancellation. But Hans posted the following just minutes ago:
Thank you for the numerous supportive comments on this topic. Several people who have experience on software or similar projects know very well some of the issues involved and have made some interesting observations in this thread.
Firstly I want to reassure anyone who is worried: the QSX project is certainly NOT abandoned! It is the most exciting and ambitious project I have worked on and I am absolutely committed and determined to finish it and make it a huge success. As soon as possible!
Yes, it is taking time. More time than I had hoped :-/ Part of this is due to increased scope. Public knowledge of what's in the works is a dangerous game... on the plus side a number of suggestions come in and these certainly contribute to build a better, more flexible and full featured product. On the other hand, more functionality means more development to be done. Which must inevitably take more time.
Other factors include: yes, a certain number of the sub-projects involved are entirely new to my skillset and experience; of course there has been a steep learning curve to climb on those, and that takes time too.
There are of course numerous other demands on my time. Sales of the existing portfolio of QRP Labs kits have been sustained at steady levels. Which is of course very welcome as it prevents any tiny mouths going hungry here at 'labs HQ... but... of course that does all take a certain amount of time to manage and support too.
Then there's also the plain old normal routine underestimation of the enormity of a project. Which is familiar to almost everyone I'm sure.
It would be well to remember that what QSX will achieve, is very high performance and features, at a low price... a combination which has never in the history of ham radio been seen before at anything like this extreme ratio (performance+features to price ratio). One should contemplate for a moment why that is, and understand that quite likely, if it was all that easy, it would have been done before already by someone. But it hasn't been, and no, it is NOT all that easy!
I am also determined not to launch something half-baked. It is crucial to me that I understand every aspect in detail and perform sufficient experiments to validate the performance and function in each area.
It is tempting to think that because this is a Softsare Defined Radio (SDR) and that therefore it can be released quickly, and any unfinished business can be later finished via firmware updates. Unfortunately this is not really the case! Not least, because there is no such thing as a completely "software defined" radio. There are various degrees of software definition... for example, Direct Digital Conversion (DDC) radios such as the Elecraft K4 and Icom IC7300 do a lot more of the work in software, than more conventional SDR like the Elecraft KX2 and KX3 and indeed the forthcoming QSX. But even DDC involves substantial amounts of actual hardware. Not only the physical knobs and buttons but also filtering, amplifier, transmit/receive switching, pre-amps... at a minimum!
Thinking you have the hardware fixed and anything else is just software, so can be finished off later after launch... is a VERY dangerous game and a painful lesson I've learned from bitter experience. There is a huge risk of something being discovered later that DOES require non-trivial hardware changes. When multiplied up to the scale of sales that QSX will likely see very quickly after launch, this could easily become a total "game over" scenario. So this is not a road I will even contemplate! Business, and life, are all about risk assessment and risk management. Neither NO risk at all, nor totally INSANE risks, are a great idea. Somewhere in the middle is what we want, after careful detailed consideration.
So... some bugs are always likely to arise. Particularly on such a complex project. It is inevitable. But we should at least start out believing that it works and is complete.
History is littered with a great many examples, even lots of recent ones, of products that are launched filled with design flaws and incomplete consideration to critical details. It's not my place to name and shame other designers and products here, but probably more than a few of you have experienced such products first hand.
I remember with QCX I took pre-orders at Friesrichshafen in June 2017 then the YOTA was in August 2017 and the launch later, 21st August. Those few months of having waiting pre-orders were pretty stressful. Good luck to Elecraft with their K4 pre-orders, they are braver than I intend to be ever again! I learned my lesson on that one. QSX... following YOTA, inevitably the news of the development became known. That has also caused a lot of stress and, is another lesson I have now learned... in future I won't be announcing any products ahead of time (not more than a week or two anyway); I will keep discussions within a small group of close friends and that will give me the benefits of motivation, moral support, and functionality/features suggestions.
On that related note, finally remember that I am a human. Not a machine. As such, I have strengths and weakness. Human failings. Emotions. Some of these one can combat and improve on. Many, not - one has to accept and know oneself and work around any suboptimal characteristics as best one can. I can certainly tell you that after much very intense work on QSX, for a long time, I did feel "burned out" on the project.
As such, in recent months, I have not pushed myself 6-feet-underground by forcing myself to only ever work on QSX. I have allowed myself the luxury of interrupting myself with work on some other projects also. But I have found that in every single one of these other projects, a substantial body of the work has contributed surreptitiously, to the development of QSX!
Where I believe it will help the QSX project, I have not been afraid to seek the assistance of others... but this is also something that is not a golden bullet solution to every problem. Quite often more manpower and faster completion of projects do not automatically go hand in hand.
the QSX project IS coming!
It will not take forever.
It will be done as right as with all my heart and soul and experience I am able to make it.
It will be worth waiting for.
It will be awesome.
It is sometimes hard to be patient waiting but, you must try.
Thanks for your support
73 Hans G0UPL