Half baked software/firmware
Posted by: Redboxfile on 08 April 2018
Further to my recent post regarding my frustrations using the Naim app for Android I came across a damning article in the June issue of PC Pro magazine written by Jon Honeyball. The article highlights the poor implementation of firmware and software support by leading manufacturers. The Naim Uniti Core was featured and also Chords Mojo as examples. It makes the point that it is so easy for companies to tarnish reputations built up over many years and that "crawling out of that hole can take years". So often these days companies release products that are not ready for market and then need firmware updates to put things right that should have been right in the first place.
In the software world it seems all to common for the purchaser to be sold something that is barely fit for purpose, with problems that should have been fixed at the beta-testing stage. Maybe that is easier to understand where software is used on a vast numbr of different computer setups, where other things can interact, but less understandable when it is firmware for a sealed box as with Uniti last year, which suspect was a result of over-eager marketing people - ridiculous because the damage to reputation risks destroying the company, and Naim certainly can’t afford to do with the next release, and I guess the streamers may be delayed, but work when they come out.
Android apps, and firmware updating software run on a variety of external computers, unfortunately potentially suffer from the same variable factors as much other software, where beta testing can only cover so many combinations and permutations...
The number of large motorcycle manufacturers who release new models with terrible fueling problems due to maps being poorly developed is amazing
Again ..this can completely tarnish their reputation and the model concerned
And a trip to a Dyno for a remap shows what can be done...but at a cost to the customer to " developed" it
Perhaps I shouldn’t say this because I am in the IT industry. It’s sad that recent years there’s real strong movement towards agile but few are really doing agile right. Temptation is to Just put something out there - minimum viable product. To claim that releases will come out often. Don’t get too hung up about bugs because they will be fixed tomorrow in the next drop and you get new release a lot quicker...
I am all for continuous improvements. But there should be a roadmap and there should be some due diligence to the minimum quality it must deliver and that quality should be based on realistic release frequency, nature of the product and user expectations. Not just minimum and nothing more. So often MVP is an excuse to churn out half baked and deficient solution. So often good strong design end up taking a backseat and end up building ever increasing debt waiting to be addressed later.
If I am right, Naim, based on how it targets the market, should be aiming to produce products that feel like well made, well engineered finished products. I doubt it’s loyal customers are those with an obsession for updating firmware for the sake of fixing issues. Again, continuous enhancements and support for future tech is great and I am all for it. And Naim’s commitment to single codebase at least per product line should be applauded because of its complexity. Please don’t forget tech debts and spikes are very expensive.
Unfortunately 'agile' is often combined with 'App' development for portable devices, and that very often means inexperienced developers who lack the necessary experience to know the right way to do things upfront and who also don't have an attitude of thorough unit testing. This then puts excessive pressure on the system testers (who, due to impossible project timescales and marketing deadlines, usually don't have all the planned time anyway), then bugs either go undetected and get through to release; or the bugs are detected and the marketing department just releases the App anyway.
Mostly it's management failure for making deadlines too short and employing developers who are simply not good enough for the job.
You only have to look at the Google play store feedback to see that there's a lot of unhappy Naim users out there. Apparently not all developers, including Naim, look at this feedback though. And that's a terrible indictment of the present system. I was very disappointed with Naim android app but I'm not so anti now that I have got used to it's many shortcomings, most of which I've managed to work around. Still testing the latest firm and software updates and all looks good at the moment, but I will reserve judgement for a few weeks. I do accept that there will be firm and software updates to address issues. I have also found Naim support very helpful. Contacting support should be the first port of call before moaning. Rich
Oddly with the change from 2.5 to 2.6, one of my android devices now works quite well.
The tablet used to work about 50% of the time (although sometimes it world refuse to work for several days on end), but the 'phone would work once when the app was initially started but then never work again once the 'phone has gone to sleep and woken up! My other 'phone won't even load the app.
Now the 'phone works reliably (albeit with some slightly 'idiosyncratic' behaviour, but still working), whereas the tablet works once when the app is initially started and then never works again after the tablet has gone to sleep and woken up!
I’ve said it before: I wonder if it would help somewhat if Naim would say “we support the Android app on the following phones / tablets and OS version(s)” (and give a short list).
Downloading new 'Frimware updates' can wipe out your saved information. My experience is if you are happy or content with your present frimware on your particular component downloading further updates will be at your peril...