Of course. Sorry about that, I didn't immediately link your mention of a bug with the thread directly. Yeah, all is clear now. Incidentally please don't think I'm just being argumentative or trying to put your thoughts down here, it's just interesting to get a perspective from the opposite side of the question and to have your own ideas and experience tested and criticised at the same time. That way you maybe see where you are currently wrong or something different and new and you become better informed yourself. Ok, the last thing I would point out is this and then I'll leave you to it.
mlordin wrote:As for Firmware updates, a majority of digital musical equipment manufacturers use downloadable firmware updates successfully to the benefits of the user communities - examples, Fractal Audio, Kemper, Line6, TC-Electronics etc, etc. Most software companies always have to consider the backwards compatibility issue when making changes but typically works pretty well - as a user you can always opt not to update your equipment in case you don´t see the benefits.
Now there are some interesting points you raise there but I would urge you to bear in mind that all of the companies you mention are dealing in heavily programmed intensive real time digital manipulation almost exclusively. Their software will be large with many thousands of lines of code performing a huge number of continuous real time mathematical tasks as well as logical ones. And the results of those complex manipulations can often be tweaked by performing the work to a different strategy which means altered programming. There is a benefit to them having ongoing development which they need to pass on to their users, the task is complex enough. As a simple example, I have a Line6 JTV59 Variax guitar. It's modelling of different guitars is performed by having a digital template for each one which is used to mathematically alter the digitised signal of the guitar's own piezo pickups in the bridge. It isn't in any way a digital version of each guitar's signal, it is more a kind of a mathematical modifier which gets imprinted on the guitar's signal, (anyone understand convolution?). From time to time better model templates are produced and these are made available to owners so the guitar can be updated from their own software control suite which you get with the guitar. That's a genuine need where things can be improved so the update route is necessary. And it's even more necessary for something like the Kemper or TC Electronic units which are much much more comprehensive than my guitar.
The GM40D and GM36, (and my Marshall JVM205H for that matter) are just not that type of beast. They are performing ultra simple logic functions like - "detect a MIDI message, interpret the couple of characters, change a few settings", or "detect a change in a control's positional value, reset the control's current position, send a message to the control to change its resistance", or "monitor the output bias conditions, detect a change, adjust the bias", alongside its most intensive task which is applying the digital effects. Even that is not at all a demanding task for the simple effects it has on offer and once established would not need to be upgraded at any time. The processors are really basic ones, (a couple of ATMegaA88 and an ATMegaA168), with a very small memory capability of 8KB/16kB flash memory, 1KB RAM, 512B EEPROM, reflecting the simplicity of their tasks. That's 3 processors, not including the digital effects on its own PCB, all to be updated separately. If you think about it in those terms maybe you can see why H&K can update them back at base but the user can't.
Now you found a bug which is very genuine, but you have to admit, it is a very obscure bug and as such is not one which would ever go on a software company's list of criticals. I'm sure it will be fixed in a future model release when the whole software package is modified to suit but it is very unlikely to be addressed as a retrospective issue in itself even if the update pathway was available, it just isn't economic for them to work that way. Bug fixing is generally done sporadically by engineers who have no other task on at the time and it's priority driven. (It's also generally disliked a lot!) The amp does not fall over in use or risk damage because of that bug, ergo, low priority.
I would think of it this way. If it was found out that your car had a brake fault which could mean that the brakes didn't apply under certain obscure circumstances the manufacturer would call that in immediately and deal with it. If it was found that the boot could pinch the cable taking power to the rear end of the car when it was closed leading to a few wiring problems after a number of years they wouldn't do a recall for that. They might put the info out to their service centres to watch out for it, they might even plain ignore it as not really important. That's what I think we are dealing with here.