After the latest update which brought so many performance issues for so many of us it seems that the option of adding a Beta version before a Sim / World update is crucial…
It is planned for this year but I guess this could even mean in Novermber or December of 2021, while in the meantime we will have pretty major updates till then…
I remember Jorg mentioning about it in the last Q& A video after a community question, if I am correct, and he didnt sound so urgent about it…( I could be wrong about that but that was the feeling I got)
However I hope I am wrong and this option will be introduced to us earlier this year because it would let us continue enjoying the sim without these hicups and save the developers from our complaints…
For instance now we would all roll back to the previous version and we wouldn’t be starving for a hotfix…
Anyhow I went and voted for it…
The process might start with third-parties getting access ahead of releases so they can align and test their own products to be in sync rather than a trail behind model.
Part of the challenge is standing up a separate infrastructure to support Beta or Preview Release testing without being co-mingled with or impacting production users on current version. That’s time and money that may not have been allocated in planning/budget. So yes, the bottom line may be that it will take time. How long is really anyone’s guess.
Yes…but is it really so difficult when I am starting the sim and the check for updates and then the update screen comes up to give me the option to proceed with the update or just continue with my current version?
I may be simpifying it but is this so difficult? We had it in Xplane 11 since I dont know when…
X-Plane and Euro Truck Simulator have beta branches, so it surely could be done for MSFS too.
I still have Xplane installed, even though I am not really using it, and there is an option to get notified for beta versions or just for the final ones. Once I open the sim I get notified, I choose to install the update or not and thats it…
I sometimes think that Asobo/Microsoft dont want to call these updates “beta” for some reason…because this would probably ruin their roadmap…but on the other hand they would have their customers more pleased.
None of the contemporary Flight Sims in the market is a Cloud based application. They are all stand alone systems on a PC, with the exception of whatever add-ons that might require server connectivity.
There is a back end dependency that requires standing up a parallel environment to service the authentication, weather, traffic, ATC, Maps & Photogrammetry streaming. That’s expensive, and takes time.
Yeah you are right…I havent thought about that…
In the last Twitch Q&A, Jorg kind of implied the beta testing branch is up for discussion again and they are discussing it. I was like, what? I thought the beta testing branch was a sure thing and was going to be deployed some time in Q1 of 2021. I remember in the fall of last year at one of the Q&A sessions, they specifically said they planned to deploy a beta testing branch. Now they are back to discussing it again and whether they need a beta testing branch 
I hope Asobo can be more specific about the beta testing branch then. They should tell us if the beta testing branch was only for 3rd party developers, or whether they planned a beta testing branch for the wider public and general MSFS player base.
Back in fall of last year, it was clear the community wanted a beta testing branch at least made up of MSFS players and not a beta testing branch only accessible by 3rd party developers.
There’s no guarantee that standing up a Beta Branch would improve the quality of releases. Part of the challenge is to properly manage and triage what is happening in Beta and ensuring that the flow and volume of bugs is material and substantive. In many ways, it almost becomes a Bug Hunt Bounty Program minus having to pay out.
At the end of the day, this is a business, and having to stand up a separate infrastructure means capital expense. Once stood up, it’s not sunk costs but a continuous expense that needs to pay back somehow.
Third parties getting access buys some indirect benefits, namely a developer Community that likely will draw more applicants if the experience is positive, as well as goodwill by existing sellers. But that’s it. It doesn’t directly sell more copies or guarantee in some way that the Product Vision’s core features get developed and delivered.
They should probably be more specific in general…
Amen Brother 
It may be expensive but they are probably losing customers, or they have their already customers disappointed for not having a constant good experience with the sim…Why should I or you be obligated to update to a version that is broken?..(they dont intend to release a broken version but thats what usually happens…)
If they manage the beta branch properly and have a good system of sorting the information correctly, it will definitely improve the quality of releases. Beta branches are nothing new to the video game industry. Almost every other complex online video game has a beta branch that is tested by their player base. This isn’t a new idea or new technology. It’s pretty much a standard out there with other large and complex video games.
And they also waste money tracking down bugs like the performance bugs for this hotfix, and then retesting for this hotfix. Without a doubt, some of their developers that could be working on code to advance the simulator and add more features, are then pulled off those tasks to help track down and fix bugs that would have been caught by a beta branch tested by the player base. And then their Q&A testers are diverted to help test the hotfix patch, when they could have been testing new 3rd party products for the market place. In other words, forward progress is halted and they spend more time backtracking and fixing stuff that would have been caught in the beta branch testing phase (if you do software development, you know that it’s much cheaper to fix things in the testing phase than if you have to fix it after the bug goes into production). That’s $$$ lost because they didn’t have a beta testing branch. If this is going to be a 10 year project, a separate infrastructure for a beta testing branch will pay itself over the 10 years, than having half of Asobo scramble to track down and do hotfixes over stuff that should have been caught during beta testing.
I’m not convinced a Beta Branch will improve matters. A good majority of Beta Testers will fall off after realizing how structured actual Beta Testing is. It’s real work, if done right. Some will stay on because they are true believers, but the majority, once they realize, hey this isn’t actually me being free to just bang on the sim in my own way, will drop off.
Then it becomes a matter of metrics in terms of returns. How many testers are left, what is the number and quality of the bugs they found, and how many can actually be fixed in this cycle of the Sprint? Or are they found but put in backlog, which means they’ll exist anyway in Prod in some way shape or form until the schedule gets to that backlog item, or priority gets shuffled? It becomes a defacto job. Some will do it willingly, but all too often, folks drop out of it because they realize how much work it really is. What you have left has to pay it forward in terms of either delivering true bugs that can increase quality IN the current cycle. That’s a big ask with unknowns. So yeah, if someone’s being asked to foot the bill for a six figure minimum Cloud infrastructure, that business case better be ironclad.
Beta testing is happening before releases
https://steamdb.info/app/1250410/depots/?branch=public
And as others have noted, even after beta, bugs will stll show up. There are 100,000’s of lines of code any changes will have unintended knock-on effects somewhere in a complex system. Its really the nature of the beast!
I think you’ve hit the nail on the head there. Lot’s of people shouting from the rooftops about implementing testing, which I agree is a valid idea, however it needs people who are going to test specifically and exactly what Asobo ask them to, and not just fire a pre release version up and treat it like advance screening tickets to a new action movie.
How Asobo plan to do it to make it effective for them, and the willingness of people to actually test what’s there to be tested and report it, is as you rightly point out, the dilemma.
Living through the X-Plane open beta last year was nigh on an exercise in losing faith in humanity.
True, but XP11 is not tied in to online services in any way, except for METARs, and their updating mechanism. MSFS is for all intents, and purposes an MMO. You don’t get to choose which version of those you want to run.
Those who have fallen foul of the updates will know what happens if you managed to get the client to run without having all the updates installed. All the online features are disabled.
If users were given the option of a client update that had no access to online features at all, or at least those dependent on a current version of the sim, I guess that could be doable.
Side effects would be no access to updated content from the marketplace, third party planes outside the marketplace when updated would likely not work correctly as they would be based on the latest version, so no bug fixes there.
But if you wanted a snapshot of your system, frozen in time as it were, that could be doable.
The question is whether they would spend the time developing that.
Well MMOs have things like persistent worlds, PvP and other aspects which would break them if players are on different versions.
MSFS is really just streaming data which doesn’t really change from release to release. Also it pushes positions of the players to the cloud but that is also just a limited set of data like it’s done on VATSIM even for players with different Flightsims.
Sure there are changes (especially new features) that require changes in their backend as well, but if you account for that (which is a bit of work but by no means impossible), you should be fine to support at least two different game versions at the same time.
Marketplace as well as community addons already have version dependency definitions so they don’t run if you don’t have at least the version they are designed for.
Basically all App developers for smartphones do this the whole time since you can’t really force users to update. And they sometimes have even more limitations for the version between App and their Cloud backend.
Companies do Canary Releases, Feature Flags etc. which are very common ways to try out new features without affecting all users.
So let’s not make it sound like it’s impossible. And the effort would surely pay out once the platform is more reliable which is by far better than people complaining about issues under every single tweet of MS.
Does the argument that optional updates are not possible (because of potential multiplayer issues) assume that everyone plays using multiplayer?
I don’t.
I realise that there are live elements to the game, such as scenery and weather but I don’t think they’re particularly dependent on version (unless the API is changing very frequently, which is pretty poor planning in my opinion) and they’re definitely not dependent on other players having parity with me. There may be times when updates are required to cope with online API changes. But these would be critical updates (coping with breaking changes in dependencies), which is what (I assume) other single-player games also provide.
What if I don’t update, therefore I don’t see options for making multiplayer features active? What I mean by this is some of these ‘blockers’ might just need a bit of thought and planning applied but I don’t see how impossible it would be.
It seems a lot of people would enjoy the ability to update at will. I would because multiplayer isn’t a thing for me. At all.