How can a good God allow suffering

I think that is correct.

And neither was I when I have said "Biological organisms are software defined lifeforms." It's just a factual statement about the phenomanon in front of us. Any ontological inference is in the mind of the reader.

Sorry, I should have made clear that 'medium' and 'environment' tend to be usec equivalently in Maturana's writing.
You will not get the essence of how he departs from standard observational reporting e.g. of 'behavior' unless you are prepared to read the references. What may be 'agreed by standard observers' is relative to their consensual domain, but may have no 'meaning' from the pov of the autopoietic lifeform.

Unless we discuss this further within a common semantic field (consensual domain) generated by the references, we will be talking past each other. For example, what we as 'standard observers' mean by 'behavior' may have little to do with the concept of 'behavior' which Maturana ascribes to 'languaging'.
Its a similar situation to trying to discuss, say, Heidegger's Sein in terms of ordinary understandings of the word 'being'. You need the whole edifice of Heideggarian neologisms to give a coherent framework.
(The von Glasersfeld comment of Maturana's 'circularity' is significant in this respect)

That was a metaphor, as I understood it. And as I explained metaphors have limits, including this one.

That is where you lose me. I don't believe that ideas are necessarily bound to certain semantic choices. In my view, one should always be able to express one's ideas in a language understandable by a ten year old, and philosophical texts ought to be translatable in other languages than the one they've been written in.

Popper says this in the following way: the complexity of one's language should always be kept to the minimum necessary to deal with the problem at hand.

The problem raised by this thread concerned the utility of the concept of 'information' and why Maturana has rejected it with statements like 'an autopoietic system ie informationally closed'. That in turn leads the more general issue of 'the observation of observation', and why Maturana confines 'observation' to 'language users'. So to talk about 'simplicity of language' is a non-starter for him as a biologist/philosopher, in the same way as to talk about 'money' as merely bunches of paper and coins, would be to an economist.

NB Further to Popper (and Einstein) comment on 'simplicity', Niels Bohr said that words applied to quantum theory were at best 'poetic', so Maturana sets no precedent, even allowing for the probabalistic nature of QM, which elludes the Falsifiability Principle

The problem with needlessly obscure language is that, when used by imposters, it can look profound while in fact being vacuous. E.g. Hegel's and Heidegger's styles hide more than they reveal, on purpose. It's a defense mechanism, a posture designed to avoid debate and scrutinity. It glitters but ain't real gold. Foucault called Derida an "intellectual terrorist" for similat reasons.

Obscure, idiosyncretic language also lends itself to many different and divergent interpretations. To a degree it's true of any language; human languages are always polysemic. But one can reduce ambiguity by using clearer language, or coumpound ambiguity by using unclear words.

I've been reading stuff by/on Maturana that I agree with. E.g.:

"Living systems are cognitive systems, and living as a process is a process of cognition. This statement is valid for all organisms, with or without a nervous system."

But I am not certain you understand these words the way I do.

I guess answering a question in one's own mind is easy - when you alone get to decide what the question is.

I understand it to mean that 'cognition' refers to the process of 'adaptation' (assimilation +accommodation) common to all lifeforms. From the systems theory pov, we are looking at 'state transitions' of an organism which define for it corresponding 'states of its environment'.

We've discussed the software-hardware metaphor at length and I have pointed out some of its limits. You're entitled to disagree but I see no point in rehashing all this again.

And I understand it to mean something like 'information management for survival', eg remembering where food sources have been located in the past, so as to revisit these areas to search for food, or being on the look out for predators in order to flee or hide from them, or using photoperiodism to trigger flowering at the most opportune time in the year.

That is incorrect. It's perfectly possible to prove a probabilist prediction wrong. It's done all the time in medecine and biology. There are statistical tools developed specifically for that purpose, eg ANOVA.

Maturana argues that 'learning' involves 'structural change', not 'information retrieval and processing'. I would use the analagy of 'structuring a keyhole to receive the key'. No keyhole - no key. Recognition is physical not what we normally call sensual. For M there are no 'sense data' That's the big leap!
(...it reminds me of Einstein's dismissal of 'time' as a fundamental parameter).

In statistics 'wrong' simply equates to failing to reach 'a specified confidence level for rejecting the null hypothesis'. It is quite common in social science, at least, for researchers to purposely or inadvertantly manipulate choice of tests and confidence levels to 'bend' the results.
(I understood that Popper had to modify his reliance on falsifiability in order to accommodate statistical subject matter but Ive not been into it in detail).

Why do all animals have these organs we call senses (ie eyes, ears etc) then? It would impossible to make sense of say, a falcon eye or a bat ultrasound echo hearing system without assuming that they capture some data that's of use to the falcon or the bat.

Yes but when you can conclude that the chances of hypothesis X being true are equal to 0.0000001, you've pretty much falsified the hypothesis. And there are no sciences that I'm aware of which can always yield the probability that a given hypothesis is true equal to zero, or equal to 1. Science rarely offers absolute certainty. Which is where Kuhn improves on Popper, with a more nuanced, less hardcut view of how hypotheses and theories get to be rejected in actual practice.

With respect to 'sense organs' , M appears to consider them as sites involved in 'the perturbation of structure'. The emphasis is not on the 'nature' of the perturbation. So what we call 'perception' he might call 'adaptation to perturbation'. The whole emphasis is on 'inner process'. There are no 'things out there' for non languaging lifeforms. They either adapt successfully to perturbations or die.
Interestingly, early theories of vision conceived of a process of 'feeling with the eyes', rather than 'receiving light signals'. For me this goes part of the way towards giving an insight into what M thinks is going on in 'structural coupling'.

Of course Maturana is not without his critics for his unconventional views,, and the only 'test' we might have of his ideas is their applicability. In that respect, he has provided a theoretical spproach behind the 'embodied cognition' movement, which seeks alternatives to the spectacular failures that have occurred in AI based on solely on 'information models'. It has also been applied to the study of social organizations.

I don't let those uncertainties bother me. Why worry about the small stuff when life is replete with them.

You haven't discussed it at all, you're still in denial.

It's not a metaphor.