Fenix block 2 delayed [31 October 2023]

[ha5mvo 261]

31 minutes ago, F737MAX said:

No sarcasm.

By using MSFS flight modelling parameters, the Fenix CFM fuel burn is (IIRC) about 5% off from the real world data. To fix that, Fenix decided to go down the route of externally modelling the engines.

mmm, it's much more than 5% on the fuel burn.

Have you ever attempted an OEP with the Fenix?

[Fiorentoni 4,420]

21 minutes ago, ha5mvo said:

mmm, it's much more than 5% on the fuel burn.

Have you ever attempted an OEP with the Fenix?

Here I have to agree with you, and even with both engines it's way more off. It depends on the length of the flight, because the cruise fuel burn is wrong. On a 5 hour flight it's 10% to 20% off in the end. Can be worked around with the simbrief fuel setting (e.g. M10), but yeah, that's what I'm mostly looking forward to to be fixed with the external engine model.

[ha5mvo 261]

23 minutes ago, Fiorentoni said:

Here I have to agree with you, and even with both engines it's way more off. It depends on the length of the flight, because the cruise fuel burn is wrong. On a 5 hour flight it's 10% to 20% off in the end. Can be worked around with the simbrief fuel setting (e.g. M10), but yeah, that's what I'm mostly looking forward to to be fixed with the external engine model.

You can work around fuel burn figures, that's the smallest problem.  What you can't work around though, is what happens when you fail an engine - this is where the real party starts!

I wonder sometimes (many times) what some of the comments here are based on.

As for timeframes, I have no issue with them taking whatever time it may take if they can fix their flight model. Developers are mostly reluctant to publish such estimates and we can see the reason why.

[F737MAX 5,738]

2 hours ago, Fiorentoni said:

On a 5 hour flight it's 10% to 20% off in the end.

Good grief. In 5 hours I'm doing 2, possibly 3 legs across Europe as I don't have the attention span to finish a...

Yes, that margin of discrepancy is too much for a high-end complex airliner.

[cavaricooper 2,615]

Curious to hear if Cupid has anything in store for Fenix customers this Valentine’s Day? Any further tidbits from Discord?

Ta!

[Jeff Nielsen 1,403]

58 minutes ago, cavaricooper said:

Curious to hear if Cupid has anything in store for Fenix customers this Valentine’s Day? Any further tidbits from Discord?

Ta!

Nope, just people complaining their communication is poor even though they do communicate...lol.  The 1-dimensional internet amazes me sometimes.

[cavaricooper 2,615]

No complaints here… just happy enthusiasm… I do really like her!

[Zimmerbz 168]

2 minutes ago, Jeff Nielsen said:

Nope, just people complaining their communication is poor even though they do communicate...lol.  The 1-dimensional internet amazes me sometimes.

I'm curious.. what have they been communicating other than it will be ready when its ready and they are in the wrapping up phase?  I hate discord so I miss a lot of info that is passed through there.  

[Jeff Nielsen 1,403]

3 minutes ago, Zimmerbz said:

I'm curious.. what have they been communicating other than it will be ready when its ready and they are in the wrapping up phase?  I hate discord so I miss a lot of info that is passed through there.  

They have to keep repeating themselves, therefore people all the sudden complain they're not communicating.  Makes no sense.  I mean, what do people expect to see in this phase?  LOL.  

They are wrapping things up, code cleaning...then I'd assume putting it in the installer, liveries, all that stuff.  Probably need to test it all as well.  It all takes time.  There's literally nothing to see until they release.

Edited February 6 by Jeff Nielsen

[Zimmerbz 168]

1 minute ago, Jeff Nielsen said:

They have to keep repeating themselves, therefore people all the sudden complain they're not communicating.  Makes no sense.  I mean, what do people expect to see in this phase?  LOL.  

They are wrapping things up, code cleaning...then I'd assume putting it in the installer, liveries, all that stuff.  Probably need to test it all as well.  It all takes time.  There's literally nothing to see until they release.

Ok.. That is pretty much what I knew as well.  I just wanted to make sure I didn't miss something in Discord.  And I agree with you, I am excited for this release but people make it sound like it is going to jump out of your monitor and be a real aircraft! 

[cavaricooper 2,615]

From Aamir, just now….
 

For whatever it's worth - it has literally nothing to do with MSFS2024, we know nothing about it - same as you. As for delivering nothing - well, yeah - we can only deliver an update when it's ready, and until then all we can do is discuss it with people that say things like "it's all about the useless fluff", but I'm afraid that is a bit misguided as I elaborated above.

But, it's fair criticism, I hope you won't mind me sharing my perspective and thought process, along with my learnings.

1. IAEs closing in on being done. We switch from (most) work on the IAEs to work on the CFMs. Our outlook at this point is that CFMs will be 3-4 months of development, less time than the IAEs, but still quite far away. At this stage, releasing just the IAEs made sense. New CFM XEM tied to new visual model in animations, sounds, lookups, LVARs, etc etc from a core level (XEM requires fiddling around with a lot of this stuff, like FM etc too).

2. CFM moves rapidly thanks to the infra being easier to work with when building a whole new engine. Revised internal timeline is 1-1.5mos. IAE internally expected as announced. However, there is a catch. Deploying it to the public means supporting both engine infrastructures at the same time, something you don't want to do - because it is literally 2 _different_ Fenixes running in two completely different environments, right from FM, all the way to visuals/sounds, to FADEC, to ECAM, to Failures, and pretty much.. all of it. So, the logical conclusion I drew here is to delay the IAEs, and add the CFMs to the package with a +1mo delay to cut out the risk of something going wrong with the above. 

Consideration:

+ Much faster release of the "package", you get both quicker than if you'd had one, then the other, as no support time in between to sustain the RTM version.
+ Simple packaging makes timeline predictions more reliable (that aged well..), derisking the potential of something serious going wrong with the IAE deployment and having to divert resourcing into fixing that if the need arises, better for long term objectives like A321/A319.
+ Less potential for mixed-infra to throw up problems to end-customer such as broken/mixed FADEC code (we checked for it, it was _fine_ - but scaling up 1000x from beta to RTM is a different ballgame)

- B2 art is not ready for CFMs yet, they'd need to finish it up.
- Obviously, finishing the CFMs is a prerequisite.

Some risk factors present, but managable for ultimately a better result with a net delay of circa 1mo.

3. Make decision to move release by 1mo to integrate and finish CFMs too.

4. Art team in the midst of finishing new CFMs, tracking OK. Then, as Dave explained some time ago (sorry - I'm not an artist, so I do not want to mischaracterise the issue - but he provided a deeper insight in public channels a couple of months ago) - they ran into an SDK  issue that meant any _new_ exports were utterly trashed, and by this I mean completely broken. Not just "not pretty". No documentation from Asobo. We are using SDK features that are clearly perhaps not in widespread use - we are also working _around_ a lot of already silly SDK issues to begin with. Export not playing ball. Report to Asobo, no comment provided. Art team begins process of figuring out what is wrong with no documentation. Cannot find it. Continue attempting to find the issue, as B2 CFMs are integrated into the new art - we cannot roll back easily. 1mo of integration work (just for CFMs) to go backwards is likely a larger step back than trying to figure out what is broken in the first place. Hindsight is 20/20 et al. But we now have unfinished CFMs, and finished IAEs, unable to actually be exported and used in sim.

 

The rest is more or less history. They attempted to fix it right up until the last day of October, where we would know if sorted, we had a week or two left to finish the CFMs. Maybe release just the IAEs and axe the CFM from the package for 7 days while they applied temp emergency patchwork "completion" or something like that. The calls were made on the ground at that point depending on what we found and how bad it was. Our end of October announcement did not have a timeline for that precise reason - we had no idea what we were dealing with. On that basis, how does one make a reasonable decision with a lack of information? But we were also ultimately _hard stuck_. We could not export anything save for the old model. 

Ultimately, mid October, we figured where the issue was - and sadly, it wasn't really anything we could influence, it would be down to the provider of the SDK to do so - so the options were wait and hope the ticket gets answered (and answered positively), followed by an undefined period of waiting for a fix to come through (or the risk that they state they aren't fixing it at all) - and then carrying on - or just working around it. So we opted to work around it, requiring a substantial rebuild of a lot of core bits of the affected artwork which had already been produced. 2.5mos later, on the 1st of Jan, it was finally then put into wrap up phase after a _heavy_ amount of undoing/redoing our work using a _tonne_ of workarounds instead of touching upon the poisoned well, so to speak, that created our export issue in the first place. This wrap up phase is longer, of course, just by nature of the sheer amount of stuff that's been rebuilt that needs to be checked over and optimised for RTM. So, it worked out okay in the grand scheme of what is effectively the only two viable options available to us - given it could have been so much worse if we rolled back, then attempted a fix. You'd have gotten a sub-par quality update after a loooong wait, my timeline had it landing around about now if we opted for rolling back, followed by another 3 months or so to get the visual update + all the extras (not to mention adding 3 _extra_ months to the A321/A319 given they are exclusively on B2 art) - or you can get what we intended for it, in the coming weeks, with, yes, a delay to A321/A319 but damage mitigated to at least not be as bad as it could have been. 

I realise looking back at it, yeah - the decision to delay the IAEs to include the CFMs caused a _sucky_ situation to turn even worse - and I'm not gonna blame luck, but really - we sit in a position where are liable to having a massive punch in the face when one of the other programs we use to build the airplane decides the SDK works differently today. Or an undocumented feature we're using is just going to.. stop working. I made my decision to delay the CFMs on the basis of the risks I saw in front of me, on my little risk matrix. What came and sucker punched me in the face wasn't even considered. But I did learn a valuable lesson from it, in that a bird in the hand is worth two in the bush. Clearly, it was poor decision making on my part having seen the upside to doing it the "faster" way, and it ended up biting us in the word not allowed big time. I take ownership of that decision, but that is my perspective on it and what I've learned out of it. 

Of course, re communication, update cadence, and a whole host of other stuff you mention - I know, we're making changes to sort that out.

 

FWIW- I like Fenix’s FIFI enough to visit and endure Discord 😉

Edited February 13 by cavaricooper

[grandfred29 180]

Where on discord you saw that?

[BrammyH 467]

1 minute ago, grandfred29 said:

Where on discord you saw that?

It's in the general chat area. If you do a search on user name "FNX_Aamir" you'll find all his posts these are there.

[redshift27 17]

Actually in #points-of-view

[BrammyH 467]

1 minute ago, redshift27 said:

Actually in #points-of-view

Ah, ok. I just search on his name to see if there is any info so I rarely notice where it was actually posted.