Thanks, Oblivion. You make some excellent points. I am of the view that the video also make good points. I explore as follows.
You are correct. That said, so is the video - from too many decades of bitter experience.
Concur. I've worked with hundreds of business software clients, and equally been client of probably hundreds of software providers. Of course you're right.
In business there are remedies. In consumer law there are also remedies (though in this context little as software licenses rarely qualify as would consumer goods). But in both cases the ultimate sanction is voting with your feet. It's the only language business understands (just as you say below).
But again, the fact you, and I, and Paradox, don't dare trot them out doesn't make them less true. I give brief reasons, and explanations (never excuses because I try to be a grownup) and take a hit, usually financial. I know I do when my service or software deliveries go wrong and the framework service agreements or whatever don't cover off the source reason.
So I have a relaxed attitude: there but for the grace of God go all of us. The only reason I've had fewer than my fair share of death march projects is I stumbled across (never was taught) various tricks to either avoid them or resurrect them. But to return to the precise point at issue, I've never solved the dilemma, in a competitive market, of whether to ship now, or ship later with a few less bugs when demand has disappeared. (In particular, as mentioned in the video, third party reliance is often a binary risk - it either doesn't crystallize or it wrecks you)
I agree with your conclusion's sentiment and humor, but on the off-chance you meant it literally I would disagree. There's a balance there. The facts set out in the video are some of the various factual scenarios that can happen, any of which can cripple or delay a project. To ship, to postpone, or to cancel are the three choices available. The decision is a business one. Sometimes the decision will be good for the business. Sometimes bad for the business. However one thing that's always true is that the decision will not please everyone.
On this point, I disagree, for two reasons.
First, Many other software companies, do the same thing once they're beyond a certain size. One company built its empire not on software (which is rarely theirs anyway, as I've personally encountered twice), but on the near-infinite ability of the global market to tolerate partial failure and literally millions of bugs, coupled with the very limited recovery options available for software licensing.
Secondly, games are qualitatively different to other software. To me one key "starter" difference is the RNG. One of my standard techniques is to set up automated "black box" regression testing. This can be done on almost any input and output "states", so long as they can be expressed in persistent data. Such techniques can help even small companies deliver consistently high quality software releases. However games use the RNG as a core element (otherwise it could be simply switched off for testing), and the RNG defeats this technique. While I can begin to imagine some partial workarounds based on seeds, I'd have to work on them for some time to demonstrate theoretical and practical feasibility, especially in multithreading and multi-platform environments; it couldn't catch everything; and it might even have to be done in a game-specific fashion, unsure. If that is right, then games simply lack some of the QA tools available to other software providers.
Finally, and this is not directed at anyone least of all yourself, but I'd love to see the demographics of this foum. If I speculated as to the volume of this thread. I'd hypothesize that the gaming market, especially PDS's market, might be above average in intelligence and education (thus articulate and assertive) and below average in age (thus enthusiastic, under-funded, and inexperienced in business). I cringe from my own memories of the sheer awfulness, naivety, and even stupidity of the letters of complaint the schoolchild version of myself habitually wrote to games companies. In those days I could have given just about anyone here a run for their money... of course I'd be banned now
