Board index FlightGear Multiplayer events

EDDF-Triangle

Virtual fly-ins, fun flies, competitions, and other group events. Find out details of upcoming events, register for competitions, or organize your own tour of a favorite location.

Re: EDDF-Triangle

Postby tdammers » Wed Jan 09, 2019 5:07 pm

AAL545 wrote in Wed Jan 09, 2019 4:30 pm:I feel responsible (in a bad way) for all this discussion regarding the 777 landing gear,
having said that I also feel it's to the benefit to improve the system (code) because obviously
a code was changed and correct me if I'm wrong but no one knows exactly what was changed!
If someone could take a look and compare the two between now and early to mid 2017 when everything was
working fine but I don't know if that's even possible.
But lets not do anything drastic and make matters worse.


Well, I think it's good to have this particular discussion - without it, I (and probably others) wouldn't have been aware of that particular bugfix, or the fact that the Seattle project is essentially dead, and that current development is apparently back on SVN.

I could easily compare the 2017 state of the 777 with the current one, and I've looked around a little bit - IIRC, the landing gear mechanism has been overhauled at some point (don't remember the details, could be that more of it was done in Nasal before, could be that the animation is now simply more elaborate), and the bugfix is fairly recent, so it's quite likely that the overhaul introduced the bug (it's very subtle, really), and you still have a version from that period around.

Long story short, I think this is a good thing, because now we can simply tell everyone who sees incorrectly tilted gear on the 777 to download that version instead of poking in the MP protocol or randomly blaming people.
tdammers
 
Posts: 233
Joined: Wed Dec 13, 2017 10:35 am
Callsign: NL256
IRC name: nl256

Re: EDDF-Triangle

Postby jomo » Wed Jan 09, 2019 5:19 pm

tdammers wrote in Wed Jan 09, 2019 9:37 am:It's free software; trying to restrict model choices is absolutely hopeless.

- We could extend FG to support multiple versions of the same aircraft to be installed, add a hash of the relevant model files (or parts of model files - it's probably enough to even just hash the list of MP props) to the MP protocol, try to pick an exact match from the locally installed models, and visually alert the user when none was found, which would then 1) explain any visual inconsistencies, and 2) prompt players to exchange information as to where to get that aircraft model (as we have been doing occasionally). So you could just have the six most popular 777 versions installed alongside each other, and FG would pick the most appropriate one for each MP player you encounter.

Well sorry: I thought I (and others) did point out that this it is not a problem with FGFS -- with or without the "launcher"! FGFS handles all models (that it finds) correctly without any problems.

The problem is if you join EVENTS were different pilots show up with different models/versions with the same ID's.
Then the Problem is that each user (e.g. an ATC) can only load one model-version at the same time into his local folders. Thus the other guy may:
- show different functionality
- or even see nothing at all.

My suggestion would be: Get an agreement between the designers to continue to develop whatever they believe is the best - but then get an agreement between designers (and FGFS) which design will be put into the FGFS-libraries! Thus everybody can fly whatever is the very best of all designs for him -- but for an event were everybody should see the model of everybody he must use the one in the FGFS-libraries (if he wants to be seen)!.
jomo / ATCjomo
ATC at EDDF Fr,Sa,Su,We from 20:00 to 24:00 CET/MEZ., see http://www.emmerich-j.de
User avatar
jomo
 
Posts: 919
Joined: Thu Feb 12, 2009 6:46 pm
Location: Mainz, Germany
Callsign: jomo jomoATC
OS: UBUNTU 18.4

Re: EDDF-Triangle

Postby wkitty42 » Wed Jan 09, 2019 5:38 pm

AAL545 wrote in Wed Jan 09, 2019 4:30 pm:a code was changed and correct me if I'm wrong but no one knows exactly what was changed!

all one has to do is look at the commit logs for the craft they are using... this is why proper repository histories are so important ;)
"You get more air close to the ground," said Angalo. "I read that in a book. You get lots of air low down, and not much when you go up."
"Why not?" said Gurder.
"Dunno. It's frightened of heights, I guess."
User avatar
wkitty42
 
Posts: 5757
Joined: Fri Feb 20, 2015 3:46 pm
Location: central NC, USA
Callsign: wk42
Version: git next
OS: Kubuntu 14.04.5

Re: EDDF-Triangle

Postby tdammers » Wed Jan 09, 2019 8:06 pm

jomo wrote in Wed Jan 09, 2019 5:19 pm:Well sorry: I thought I (and others) did point out that this it is not a problem with FGFS -- with or without the "launcher"! FGFS handles all models (that it finds) correctly without any problems.


Sure. It does, as long as the locally installed models for a given model ID are compatible on all clients.

But that is kind of the problem: just having the same model ID doesn't guarantee compatibility, even though in practice, things *are* compatible more often than not, simply because aircraft developers don't actually break MP compatibility a lot, and because there aren't a lot of independent implementations of the same aircraft model with the same model ID out there - even when two developers have done the same aircraft type, the model ID's often still don't match exactly, and both versions can happily live alongside one another.

jomo wrote in Wed Jan 09, 2019 5:19 pm:My suggestion would be: Get an agreement between the designers to continue to develop whatever they believe is the best - but then get an agreement between designers (and FGFS) which design will be put into the FGFS-libraries! Thus everybody can fly whatever is the very best of all designs for him -- but for an event were everybody should see the model of everybody he must use the one in the FGFS-libraries (if he wants to be seen)!


That would be one option - but "the designers" is a fairly amorphous group of random people scattered across the internet, and it's not like you can contractually bind anyone to anything. There simply isn't any leverage to police the ecosystem like that.

What you can do is publish lists of aircraft deemed suitable for a given event. This would work in principle, and can be done in a number of ways, e.g.:

- Publishing a custom hangar that contains only the "blessed" models
- Just saying "please use an aircraft model from the default hangar"
- Putting a list on a web page somewhere, with download locations
- Zipping up a directory with all the "blessed" models in it, and offering that as a download
- Publishing a git repo that has the correct versions of all the "blessed" models

In fact, this wouldn't even be super difficult; fgfs can take aircraft search paths on the command line, so you could just have a separate aircraft directory for a specific event, and pass that to fgfs instead of your normal one when you're about to fly at the event. In principle, this doesn't require any changes to FG itself, it's entirely a social convention. Better yet, you don't need to "make an agreement" with anyone - just publish the list of recommended aircraft somewhere, and link to that list in the event announcement. You can't force people to follow your advice, but it will make troubleshooting much easier, and if people willingly use a different model, and it doesn't work, then at least there's no need to start a discussion or get angry at anyone. You want to look good on video - use an aircraft model that works. Simple as that. And it would also remove the problem of negotiating versions while in the middle of a busy ATC session. Model not looking the way it should? Point people at the list and be done with it. Also, because it is a social solution rather than a technical one, it still allows people to use updated versions, as long as they are compatible with what you expect; and because the expectation is unambiguously documented, any dispute about it can be resolved easily by simply comparing the two versions.

One actual problem with this however is that you can't have multiple versions of the same model ID installed at once, and select between them, so you have to have separate aircraft directories to pick from for different situations (e.g., the event might tell me to use the CRJ700 from the default hangar, but I have a couple local fixes that I really like, and I don't want to delete my development copy just to fly the event); but FG doesn't follow symlinks, so aircraft that you want to have in both (or more) locations have to be actual copies, which is a bit inconvenient.

Anyway, my point was that it would be nice if FG handled this more conveniently itself, by making the contract of a given model explicit, and having some sort of mechanism to negotiate it automatically, which would make the whole situation much easier to deal with - but even without any changes, social solutions that are fairly simple and easy to implement are possible.
tdammers
 
Posts: 233
Joined: Wed Dec 13, 2017 10:35 am
Callsign: NL256
IRC name: nl256

Re: EDDF-Triangle

Postby AAL545 » Wed Jan 09, 2019 8:11 pm

Okay I'm confused, how many sources of the 777 are there anyway?
I started with the 777 from http://mirrors.ibiblio.org/flightgear/ftp/Aircraft/ until the gear issue arrived, then I switched to the 777 from http://www.github.com/FGMEMBERS/777.git
and stayed with that version up until now (still gear issue) and now there is the gitlab one which I haven't tried yet.
As far as more up to date there is some confusion on my part as well, the 777 from http://mirrors.ibiblio.org/flightgear/ftp/Aircraft/ was last updated October 22 2018, (according to the web site)
the 777 from http://www.github.com/FGMEMBERS/777.git was last updated December 5th 2018 (side note; I prefer this plane because it has 2500 ft call out and low air speed warning)



I could easily compare the 2017 state of the 777 with the current one, and I've looked around a little bit - IIRC, the landing gear mechanism has been overhauled at some point (don't remember the details, could be that more of it was done in Nasal before, could be that the animation is now simply more elaborate), and the bugfix is fairly recent, so it's quite likely that the overhaul introduced the bug (it's very subtle, really), and you still have a version from that period around.

Sadly no because of a system crash!



Then the Problem is that each user (e.g. an ATC) can only load one model-version at the same time into his local folders. Thus the other guy may:
- show different functionality
- or even see nothing at all.

This is something new to me, so to rule out if it is in fact a gear issue is to park along side someone who has the some model, from the same source and check that way!



Anyway, my point was that it would be nice if FG handled this more conveniently itself, by making the contract of a given model explicit, and having some sort of mechanism to negotiate it automatically, which would make the whole situation much easier to deal with - but even without any changes, social solutions that are fairly simple and easy to implement are possible.

This was my point a while ago already that FG needs to do this if we/developers want to turn this into a state of the art system,
I'm familiar enough with software to realize that this isn't going to happen if we don't eliminate the miss matches, in other words overloaded with bug fixes.
I while ago someone mentioned doing a merge, so it's okay to have more developers but they need to move up the property tree (sorry if not wording it correctly)
and in doing so if the code doesn't meet the requirement it doesn't merge, its that simple.
This not only has to do with the gear issue, the 777 is wonderful plane to fly but it has had issues and still does.
Naming; (might as well mention it here) when I pull my throttle back too late when landing the thrust reverser won't engage.



AAL4955
Last edited by AAL545 on Wed Jan 09, 2019 8:37 pm, edited 1 time in total.
AAL545
 
Posts: 193
Joined: Tue Aug 08, 2017 4:14 am

Re: EDDF-Triangle

Postby wkitty42 » Wed Jan 09, 2019 8:35 pm

if i'm reading this correctly, yes... and which ever one of you has the problem that the other sees, indicates who needs to update their craft to the one with the needed patch...
"You get more air close to the ground," said Angalo. "I read that in a book. You get lots of air low down, and not much when you go up."
"Why not?" said Gurder.
"Dunno. It's frightened of heights, I guess."
User avatar
wkitty42
 
Posts: 5757
Joined: Fri Feb 20, 2015 3:46 pm
Location: central NC, USA
Callsign: wk42
Version: git next
OS: Kubuntu 14.04.5

Re: EDDF-Triangle

Postby legoboyvdlp » Wed Jan 09, 2019 9:55 pm

AAL545 wrote in Wed Jan 09, 2019 8:11 pm:Okay I'm confused, how many sources of the 777 are there anyway?
I started with the 777 from http://mirrors.ibiblio.org/flightgear/ftp/Aircraft/ until the gear issue arrived, then I switched to the 777 from http://www.github.com/FGMEMBERS/777.git
and stayed with that version up until now (still gear issue) and now there is the gitlab one which I haven't tried yet.
As far as more up to date there is some confusion on my part as well, the 777 from http://mirrors.ibiblio.org/flightgear/ftp/Aircraft/ was last updated October 22 2018, (according to the web site)
the 777 from http://www.github.com/FGMEMBERS/777.git was last updated December 5th 2018 (side note; I prefer this plane because it has 2500 ft call out and low air speed warning)





I suggest using http://mirrors.ibiblio.org/flightgear/f ... nk/777.zip

This is the latest version including all fixes from FGMEMBERS (low speed warning as well) as well as a few other improvements.

The ibiblio one (not Aircraft-trunk but just aircraft) is a stable version as far as I know, not yet updates from fgaddon.

Gitlab is the most outdated, followed by ibiblio-Aircraft followed by FGMEMBERS followed by ibiblio-Aircraft-trunk.
User avatar
legoboyvdlp
 
Posts: 7164
Joined: Sat Jul 26, 2014 1:28 am
Callsign: YV-LEGO
Version: 2018.3.1
OS: Windows 10 HP

Re: EDDF-Triangle

Postby jomo » Thu Jan 10, 2019 10:08 am

Same models showing different

Sorry gentleman: There now are so many long texts with many opposing technical facts/fakes that I cannot understand nor agree/disagree anymore to any complete one - so I will not spend more time to read and/or comment those any more - prior to getting some basics straightened out! Can we agree on:

1.) -- Whenever you look from your 777 to another 777 on your screen - you will always see your one and only one model you loaded onto your local/private disk -- controlled via MP from the remote pilot!
2.) -- those two same models on your screen may look different if the remote controls are different! If e.g. the wheel file-structure (and thus the possible wheel-controls) in the model controlled by the remote user-model is different - then the wheels may show ok on one screen-picture - but wrong in the other!

3a.) -- Such problems cannot be solved by the FGFS-design because how shall the FGFS-Program know that the same control-signals should be interpreted different for the unique model on your disk???
3b.) -- That means we have to find a way to display many different versions of the same model on one screen - even if several pilots do use different model-versions (always displayed on YOUR screen as the one and only one model on your disk -- controlled by different pilots). !
3c.) -- That could only be solved by FGFS if giving different version-Identifiers to 777's, depending on the different designer groups! BUT that would mean that everybody would need to install all 777-versions from all designer-groups -- if he wants to see the other 777's in your area (e.g. during events etc.)!

That is why I wrote before that we may have as many versions of e.g. a 777 -- BUT at least for events we need one (and only one) "outside-version" of the many versions in one central library for events! And that is something the designer-groups must organize! Then this (let's say 777standard) could be used by all designers that at least the outside parts/controls may be kept similar - while the internal functions of the unique models may be different - but who cares during events as long as the external controlled models look all the same!
--> Remember: With how the model is acting depends on the model on the other guy's disk - while the look of the model depends on the MPsignals sent from that other guy to your model on your local disk!)

Sorry dear designers - but I see the solution ONLY on your side: You may be designing as many "best of" as you want in as many as you want libraries - but the external views must stay the same! Especially as a poor ATC having often different models 777 visiting EDDF - I will NOT change my local model libraries each time a different 777 from a different designer comes in.

And for pilots: Sorry - but if you want to see all versions of a 777 (i.e. during an event) you must decide on one version to store onto your local disk - if with that version you cannot see all versions that you want to see - DO ONLY USE MODELS THAT PROVIDE THAT POSSIBILITY!

Can that be a compromise?
jomo / ATCjomo
ATC at EDDF Fr,Sa,Su,We from 20:00 to 24:00 CET/MEZ., see http://www.emmerich-j.de
User avatar
jomo
 
Posts: 919
Joined: Thu Feb 12, 2009 6:46 pm
Location: Mainz, Germany
Callsign: jomo jomoATC
OS: UBUNTU 18.4

Re: EDDF-Triangle

Postby tdammers » Thu Jan 10, 2019 12:52 pm

Forget the whole property stuff; all those problems you mention of people not using compatible aircraft is not what caused the gear tilt problem on the 777. These issues do exist, and things have gone wrong because of them, but THIS IS NOT ONE OF THOSE CASES.

The 777 gear tilt issue is simply a blatant bug in the 777 model and has absolutely nothing to do with what gets transmitted and what doesn't.

Therefore, the cure doesn't require any negotiations. If you have a broken 777 installed locally (e.g. the one from fgmembers, or from the abandoned seattle project github), then you'll see the gear tilted wrong; if you have a recent one (from one of the links lego posted), then you will see the gear correctly.
tdammers
 
Posts: 233
Joined: Wed Dec 13, 2017 10:35 am
Callsign: NL256
IRC name: nl256

Re: EDDF-Triangle

Postby tdammers » Thu Jan 10, 2019 1:17 pm

jomo wrote in Thu Jan 10, 2019 10:08 am:Sorry gentleman: There now are so many long texts with many opposing technical facts/fakes that I cannot understand nor agree/disagree anymore to any complete one - so I will not spend more time to read and/or comment those any more - prior to getting some basics straightened out!


It's a bit difficult to have technical disputes with people who make technical claims but don't have the required technical understanding.

It's also difficult to try to explain things to people who interpret the situation as a dispute.
tdammers
 
Posts: 233
Joined: Wed Dec 13, 2017 10:35 am
Callsign: NL256
IRC name: nl256

Re: EDDF-Triangle

Postby wkitty42 » Thu Jan 10, 2019 4:43 pm

jomo wrote in Thu Jan 10, 2019 10:08 am:Can we agree on:

1.) -- Whenever you look from your 777 to another 777 on your screen - you will always see your one and only one model you loaded onto your local/private disk -- controlled via MP from the remote pilot!

is this assuming that a lower resolution model is not being used due to MP/AI LOD settings? the new LOD system (being worked on in 2018.4.0) uses low resolution models when an AI/MP craft is X distance from your viewport (or pixels in size)... it uses a high resolution models when the AI/MP craft is within Y distance of your viewport (or pixels in size)... the low resolution models generally would not employ all the nasal and xml to do things like make the wheels go up and down... lights blinking? sure but you might not see wheel/flap/aileron/rudder movement until the high resolution model is used... the high resolution model is the model you have installed with all the features... the low resolution models are taken from fgdata OR from the model you have installed IF it has a low resolution model with it for AI/MP use...

in other words, if you are flying wingman and are say 500 meters separate, you may be looking at the high resolution model for their craft... if you move out to 1000 meters separation, you may be seeing the low resolution model... this depends on your AI/MP LOD settings... currently i'm playing with setting the transition line at 1852m (1.0nm) so craft within that separation distance are high models... further out than that are low models...

i bring this up because this discussion about what is seen and when it is seen will likely come up again and again as the sim grows and changes... the MP protocol has already gone through some changes for the better... now the craft need to catch up and start using the new format so more eye-candy capabilities of MP controlled models can enjoyed by all...
"You get more air close to the ground," said Angalo. "I read that in a book. You get lots of air low down, and not much when you go up."
"Why not?" said Gurder.
"Dunno. It's frightened of heights, I guess."
User avatar
wkitty42
 
Posts: 5757
Joined: Fri Feb 20, 2015 3:46 pm
Location: central NC, USA
Callsign: wk42
Version: git next
OS: Kubuntu 14.04.5

Re: EDDF-Triangle

Postby wkitty42 » Thu Jan 10, 2019 4:44 pm

tdammers wrote in Thu Jan 10, 2019 1:17 pm:It's a bit difficult to have technical disputes with people who make technical claims but don't have the required technical understanding.

It's also difficult to try to explain things to people who interpret the situation as a dispute.

both of those statements are very true... see my post above about the LOD system that many users likely don't even know about ;)
"You get more air close to the ground," said Angalo. "I read that in a book. You get lots of air low down, and not much when you go up."
"Why not?" said Gurder.
"Dunno. It's frightened of heights, I guess."
User avatar
wkitty42
 
Posts: 5757
Joined: Fri Feb 20, 2015 3:46 pm
Location: central NC, USA
Callsign: wk42
Version: git next
OS: Kubuntu 14.04.5

Re: EDDF-Triangle

Postby AAL545 » Thu Jan 10, 2019 9:39 pm

is this assuming that a lower resolution model is not being used due to MP/AI LOD settings? the new LOD system (being worked on in 2018.4.0) uses low resolution models when an AI/MP craft is X distance from your viewport (or pixels in size)... it uses a high resolution models when the AI/MP craft is within Y distance of your viewport (or pixels in size)... the low resolution models generally would not employ all the nasal and xml to do things like make the wheels go up and down... lights blinking? sure but you might not see wheel/flap/aileron/rudder movement until the high resolution model is used... the high resolution model is the model you have installed with all the features... the low resolution models are taken from fgdata OR from the model you have installed IF it has a low resolution model with it for AI/MP use...


I have noticed the LOD option but haven't played around with it over MP because I don't understand it enough,
if you find a setting that works please post it.

Regarding the technical side of all of this, most of it is over my head and I don't take offence when a comment is made regarding people here commenting on technical issues when they don't understand.
Having said that I firmly believe the FG developers need to illiminate the possibility to download out of date aircraft, at least over MP. ( I believe this has been mentioned)
I've thought of this for a while now and that is, how do other Sim's manage their systems?
Xplane, FSX etc etc
In other words it's always good to see how the competitors are doing.


AAL4955
AAL545
 
Posts: 193
Joined: Tue Aug 08, 2017 4:14 am

Re: EDDF-Triangle

Postby wkitty42 » Thu Jan 10, 2019 9:48 pm

AAL545 wrote in Thu Jan 10, 2019 9:39 pm:I have noticed the LOD option but haven't played around with it over MP because I don't understand it enough,
if you find a setting that works please post it.

it won't be a "one size fits all" thing... the part for the AI/MP craft still needs some more love, though... i've been trying to understand it but haven't had a breakthrough yet :?
"You get more air close to the ground," said Angalo. "I read that in a book. You get lots of air low down, and not much when you go up."
"Why not?" said Gurder.
"Dunno. It's frightened of heights, I guess."
User avatar
wkitty42
 
Posts: 5757
Joined: Fri Feb 20, 2015 3:46 pm
Location: central NC, USA
Callsign: wk42
Version: git next
OS: Kubuntu 14.04.5

Re: EDDF-Triangle

Postby wkitty42 » Thu Jan 10, 2019 9:52 pm

AAL545 wrote in Thu Jan 10, 2019 9:39 pm:Having said that I firmly believe the FG developers need to illiminate the possibility to download out of date aircraft, at least over MP. ( I believe this has been mentioned)

out of data craft could cause FG to crash depending on their code... the whole reason craft are out of date is because they are not keeping up with FG development... steps are generally taken on both sides, sim dev and craft dev, to provide a transition period for new features to be introduced and stabilized... after that, the old methods may be removed at any time since they would be deprecated...
"You get more air close to the ground," said Angalo. "I read that in a book. You get lots of air low down, and not much when you go up."
"Why not?" said Gurder.
"Dunno. It's frightened of heights, I guess."
User avatar
wkitty42
 
Posts: 5757
Joined: Fri Feb 20, 2015 3:46 pm
Location: central NC, USA
Callsign: wk42
Version: git next
OS: Kubuntu 14.04.5

PreviousNext

Return to Multiplayer events

Who is online

Users browsing this forum: No registered users and 1 guest