You make some fair points, they have improved but not well enough IMO. Take the Canada WU, ok that may not be all Asobo but it still shows a clear lack of Quality Control from whoever is in charge - but hey it’s free and no one’s allowed to complain about freeware right (!?).
So whilst it’s all well and good that some things have improved, they still haven’t improved in paying attention to continuous issues & bugs that should’ve been dealt with long ago. That’s still a part of QC for me, not ignoring the past and fixing what they broke in the first place.
I can try to answer that out of my perspective as being a QA/Test Manager - there are many aspects which make it hard in case of MSFS to improve and test the actual quality and possible influences on their communication:
It’s not Asobos code Asobo has to maintain. Nobody of the team created the basis of MSFS from scratch as it’s FSX code they were left alone with by Microsoft. Sure they got the documentation ACES provided back in 2006 but that’s it as MS closed ACES. So, I don’t know how the documentation looks like, but if it’s similar to the documentation I see on legacy systems of this period it’s simply not enough to fully understand what actually is going on. This also explains the absence of some features from FSX: they removed certain troublesome features like the helicopter flight model to start this stuff from scratch instead which is a good move as they got full control and understanding in this case
MSFS is a pretty large beast with many relations to systems out of control from MS or Asobo. They can locally test things using mocks of these services and also can test inside of a capsuled environment which likely is also attached to other test systems. MSFS relies heavily on the connections stability so latencies or issues on certain connections will likely lead to issues - anyway they cannot test this on all ISPs around the globe and also only mock potential connection issues.
I cannot stess this enough but the community does not define a priority on features or fixes - they only contribute here but it’s decided by MS/Xbox business devision in first place. The community may be able to influence business but if they see something more worth to be done it’s done first.
Asobos team is fairly small and not experts across the board - they have to rely on external experts at some points which are fairly expensive. They cannot just order these guys as MS has to allow them to by providing this money. This money is planned for each quarter of a year so when budget is out they have to wait for help to the next quarter.
The persons on the Q&A sessions are not devs or QAs by themself, not even Test Managers. They can say something about planned features but don’t have a detailed view on bugs. Jörg Neumann is the project lead on MS side and also the single point of contact to MS/Xbox, Sebastian Wloch is Asobos CEO. The only one technically directly involved is Martial Bossard as being one of Asobos lead developers - however he seems to be only part of a subdivision internally as he cannot provide much information in general. He probably is part of some weekly tech lead meeting in which the different leads exchange information.
Restrictions - like everyone working for some different company in a project, Asobo employees likely had to sign an NDA restricting them what to say in public and defining which information can be passed through the public. The only guy one can see as being a spokesman from MS is Jörg Neumann. I for myself would not be allowed to talk in detail public about my tasks and actual work I do in my current project besides some pretty generic things, so I can imagine it’s the same thing in their case as Asobo sits in France and has to stick to the European law.
They actually do, but not when it comes to their mindset but technically. As already mentioned above development and test environments never reflect the productive/live environments. Asobo devs and QAs can never mock all possible hardware and software combinations around and they are working on systems which are basically just running Windows, needed drivers and MSFS. On Steam DB you get a glimpse on which environments are actually in place for testing purpose at least on Steam:
Whenever are things not be able to be recreated on dev and test systems Asobo devs have to go deep into analysis of the foreign FSX code basis which is millions of lines of source code.
Still they are doing a great job - new (core) features does not mean they left bugs sitting unattended, on some bugs it may just not worth to fix them alone as they may be related to some legacy code which will anyway be done from scratch (like on traffic related stuff they said it has to be basically redone so maintenance is set to minimum. To redo this elephant however knowledge has to be build and a plan has to be formed). The last SUs showed to me that they are starting to keep up with the issues as most people which had regular CTDs don’t see them anymore and performance increases.
It’s a long term project and I think Asobos goal will be to replace most FSX code on the long run to make this Sim finally their Sim which uses modern standards and it’s great to see MS to involve other companies to still bring also fresh stuff to the Sim regularly.
Nah. It’s not as clear as you seem to think. Trust me when saying that internal testing and outcome on production may differ drastically - all it needs is a differently configured server, something not deployed correctly to the production environment (like deployment of the wrong version) or something which went havoc on a last minute build. It’s stuff that should not happen but the ugly truth is: it does everywhere - on every game, every OS, every driver and every piece of software. If you search long enough you will always find oddities, but you cannot search infinitely with limited resources.
And please give us Logbook editing. I can live with some of the faults (for a while) if I can delete the flights affected by them. If I have to live with them forever I’m less likely to put a bunch of hours into this half-baked thing.
