Board index Other Hangar talk

We'll soon need a MUCH better terrain engine :-)

Talk about (almost) anything, as long as it is no serious FlightGear talk and does not fit in the other subforums.
Forum rules
Please refrain from discussing politics.

Re: We'll soon need a MUCH better terrain engine :-)

Postby vnts » Fri Dec 10, 2021 3:41 pm

V12 wrote in Fri Dec 10, 2021 8:45 am:Why not ? P3D is on older technology than MSFS2020, but can do this :

Yep. MS-FS is only DX11 - an openworld engine a game studio created for a driving sim, plugged into FS-X. Real-time reflections have been done in non-flight games for a long time, and flight games - it involves rendering the entire scene again (maybe with multiple renders to cover areas out of view but reflected in several surfaces), so it's normally very slow - and doesn't cover cloud shadows or tree shadows. It's sort of like how doing shadows using one technique for everything the Rembrandt way (standard game way) is very slow.

The compositor/ALS does the shadows, but much faster, and using multiple techniques. IIRC Icecold had trouble even getting the compositor object shadows to render fast initially, as shadows just eat up performance.

The plan with the new tech Icecold is working on is to do everything very fast and in a single pass. This will be used to create realtime environment maps - which can be used for environment reflection on plane surfaces, and even in water.

Flightgear already does real-time environment maps in the new pipeline - but fast, much like how it does shadows fast. The colours reflecting off the plane in this SOTM screenshot [1] are real-time :)
TheEagle wrote in Sun Nov 14, 2021 10:12 pm:Yeah … either a better terrain engine, or a 5000 MB/s superfast SSD

Yeah bandwidth is also a factor - more internet bandwidth, rather than hard-drive i/o - MSFS reached the max stated bandwidth of 5 MB/s at slow subsonic speeds at release e.g. [link] as the approach is bandwidth heavy. With less bandwidth the scenery quality drops.

StuartB said the WS3 terrain can keep up at hypersonic speeds. WS30 has a LoD system, which can render all the way up into space.
merspieler wrote in Fri Dec 10, 2021 1:38 pm:Tho for ray tracing to become really popular, we need to see a few more GPU generations with it... not only from NVidia but also AMD and Intel...
Because of performance impact and pricing of performant cards (even at MSRP)

Yep - in some ways RTX has become the new 4k. 4k was a convenient marketing feature for GPU manufacturers, and also monitor makers who can just divide the LCD screen into more pixels. 4k mostly dumps 3d performance into a blackhole (mostly), with more of the same plain boring pixels. 4k is more useful for 2d though. Games meanwhile are mostly driven by console performance which lagged behind PC last generation. So there was lots of GPU performance available but no game features to sell new PC GPUs. 4k was convenient. RTX is similar - it's possible to add raytracing to simple graphics. RTX does make lighting look softer, but it doesn't fix other things. Once RTX marketing started, GPU manufacturers forgot 4k existed :mrgreen: and the talk was back to 1080p as if nothing happened - unfortunate for 2/3k screen users who can't just run at half resolution to get 1080p, as they get blur.

RTX is another of those one-technique-fits all games industry things right now - it's more useful for games where there's lots of lights/explosions everywhere, and scenery geometry that's changing a lot. Probably helps to make the floor a bit wet and avoid non-RTX lighting techniques, so switching it on is noticeable. If one technique is used for everything, it can eat performance so you lose visuals elsewhere. For flight sims and other applications that have static geometry and fixed lights that don't change much, lightmaps are better than real-time raytracing could hope to be. For openworld applications that have the sun/moon as light sources there are other techniques including real-time environment maps, spherical harmonics/light probes etc. There will probably be some interesting techniques in future using the computational power of the special cores that do raytracing.

Kind regards
vnts
 
Posts: 409
Joined: Thu Apr 02, 2015 1:29 am

Re: We'll soon need a MUCH better terrain engine :-)

Postby MariuszXC » Fri Dec 10, 2021 6:30 pm

Ysop wrote in Fri Dec 10, 2021 11:51 am:Additionally the time setting is around noon with clear sky. A large pond in a mountainous area will never(!) have zero wind across the complete surface in these conditions.

Yes, anabatic and katabatic wind is a real enough phenomena. It may make life difficult at times...

And, returning to what I wrote in a previous post "we have effects where it matters" there is one effect which I think we do not have which we probably should have ('should' - i.e. it matters enough to have it). It is a reflection of the sun in railways. Having this effect would greatly help in distinguishing between a road and a railroad track, which in turn would help in VFR navigation exercises.

I wonder if it would be possible to reuse existing code responsible for sun reflection in water surface to get this effect.. Probably with some tweaking, as railroads are often at some slope, and also reflected light has a different tint to it.
INOP
MariuszXC
 
Posts: 1061
Joined: Tue May 18, 2021 5:38 pm
Location: Europe
Callsign: SP-MRM
Version: 2020.4
OS: Ubuntu 16.04

Re: We'll soon need a MUCH better terrain engine :-)

Postby Thorsten » Fri Dec 10, 2021 6:48 pm

It is a reflection of the sun in railways.


I suspect this is actually dead easy by tuning the existing effects - just nobody seems to have looked at it.
Thorsten
 
Posts: 12490
Joined: Mon Nov 02, 2009 9:33 am

Re: We'll soon need a MUCH better terrain engine :-)

Postby vnts » Sat Dec 11, 2021 7:44 am

From looking at WS3 road shaders/bugs recently I recall that:

- The model-shader handles it, and at ultra, it's a special model shader with moving ground vehicles - so every single effect present in the model shader can be used. That includes things like normal maps - or even dirt maps IIRC which can have maps for several types of dirt with different colours.

- The textures for roads and railways are placeholder and in data/Texures/osm2city/ - roads.png and roads_LM.png. If more types of maps are added they can be placed there.

- WS2 roads were defined in data/Materials/base/materials-base.xml . The properties can be set there under materials OSM_roads, OSM_railways as well as the roads/railways for generic WS2 roads. The new WS3 roads are also there, but these are WiP with some issues including being rendered by another effect right now, but specular properties can be added.

- I started work on a wiki page which lists where visuals seen in-sim can be changed : https://wiki.flightgear.org/Howto:Impro ... f_interest
TheEagle wrote in Thu Dec 09, 2021 11:38 pm:we don't even have real reflections - when you set something to full metallicness in Blender, zero roughness, all you get in FG is a little shiny point on the object … while in that video, even the buildings are visible on the water …

FG has environment maps - these are what does the reflections for surfaces that have a fresnel component - a mirror like component. There is also noise patterns to disrupt it so shiny aircraft exterior surfaces aren't too mirror like - tweaking is a matter of changing a few numbers.

MariuszXC wrote in Fri Dec 10, 2021 11:38 am:I think we will never win with commercial products in the 'eye candy' department.
If you look closely on the image from P3D (a pretty image for sure) [..]

The P3D image has a incredibly costly reflection, but is otherwise terrible - the rock textures are extremely low res even from far away e.g. FG can look like [img], [img], and detail of rock in FG doesn't relent even with the camera a few cm away [img]. The haze saves the image looking even worse - the haze is maybe a bit bluish for a dusty bare environment with no cities or plants that can release aerosol. Slow real-time reflections are sort of like game companies with plain graphics, or last-gen console graphics, liking 4k or RTX when it comes to selling their games on PCs - it occupies faster GPUs.

That image may not even be default P3D, but using art addons - in which case P3D is even worse. P3D is dated and generally looks worse, compared to FG in developed areas.

MariuszXC wrote in Fri Dec 10, 2021 11:38 am:One aspect is lack of manpower [..]

FG has the tech to look better than FS-X derived commercial sims like P3D even with paid art addons [1], and can even match X-P (which is more modern than P3D). Terrain can exceed MS-FS. FG also has the tools to massively amplify contributions. But historically FG contributors mostly think in terms of commercial flight game/sim workflows - and just look to build new terrain for their area or maybe add some landmarks, rather than explore the tech. For example, there's a steady stream of newcomers in the scenery forum who immediately try to build new scenery to improve visuals in their area. As a result, the very few people whose regional definitions content contributions have reached FGData in recent years are from CPU/GPU programming first - who often have to learn GIMP as they go along - rather than people who contributed to 3d models/textures/aircraft before contributing to scenery content/art (there are some new regional definitions projects in development, mostly from newcomers, and maybe as a result of improved entry barriers).

As for the new MS sim, that trailer is made up of very quick glimpses trying to build MS-FS hype in their gaming audience.
It's a bit like the first impressions of the P3D image - some of it doesn't hold up after the shiny water effect or other eye catching effect is discounted, or if shots were taken closer to scenery , or there was less blur/bloom/lens effects. Other aspects are really nice.
1). It's partly blur / lens effects drawing the eye or clouding the scenery. 2). Some the shots have a prominent water effect (which is easy to do and old tech) 3). Most of the shots are of scenery objects like buildings or trees with the ground covered as MS-FS uses fairly terrible photo scenery. 4). MS-FS does have really nice buildings / roads with textures, and their tree effect looks really nice at certain ranges/angles, but is low density / and can look like muffins arranged on a tray from distance (IIRC at release). 5). The cloud effect is just plugin in an available gaming plugin (tru sky like FS-W IIRC), and have limited types of cloud which is looks like porridge with candy floss stuck to it, and turns a solid white when going through it. Cloud lighting is also limited. MS-FS does have it's visual strengths, but it's also lacking in certain aspects.

Given the hardware available, the DX-11 engine isn't impressive or optimised. Mostly it's art content compared to commercial products like X-P, or non-commercial sims like FG. See the last part of the MS-FS thread for the non-trailer impressions [link], the first part was after the hype built by the trailers. Wayne was able to match MS-FS terrain in his area with hand-tweaked photo-scenery in FG, which still has a limited quality ceiling compared to WS2 in areas with decent landcover data.

Kind regards
vnts
 
Posts: 409
Joined: Thu Apr 02, 2015 1:29 am

Re: We'll soon need a MUCH better terrain engine :-)

Postby V12 » Sat Dec 11, 2021 9:37 am

Again tons of halfthruths about FSX and P3D :D
1. LOD system - 16 years ago FSX had very effective LOD system and user can fly from EGLL to KJFK with memory consumption no more than 2.5 GB. When I used FG, my memory consumption was 9 GB and more
2. P3D has incredibly expensive reflections - hmmm, why I did not observed incredibly fps drop when I activate reflections ?
3. Art addons - I use only ORBX Global, no EnvTex, Tomato or something other
4. Low texture resolution - hmmm, I do not observe it, for me, it is OK. And still better than synthetic looking FG procedural textures with strange effects. P3D / FSX with ORBX have more than 9 GB base textures. How much textures are in FG ?
5. "... on good developed areas" - P3D with ORBX looks reasonable good around the all world. If You will check regions covered with ORBX TrueEarth sceneries, result is incredible good, MSFS is even better :

https://forum.orbxdirect.com/topic/2086 ... nt=1770356

FG has 2 big advantages - grass overlay and smooth snow effect based on weather.

WS3 scenery demo is still only demo, it is too early to make any conclusions.
Fly high, fly fast - fly Concorde !
V12
 
Posts: 2757
Joined: Thu Jan 12, 2017 5:27 pm
Location: LZIB
Callsign: BAWV12

Re: We'll soon need a MUCH better terrain engine :-)

Postby Thorsten » Sat Dec 11, 2021 10:38 am

Again tons of halfthruths about FSX and P3D :D


Again, there is some inherent unfairness in the comparison between a multi-million dollar effort by a multi-billion dollar company and a volunteer effort. That would be they key phrase: P3D / FSX with ORBX have more than 9 GB base textures. How much textures are in FG ?

The ability to simply buy aerial images aplenty and pay texture artists to make something of them.

And still better than synthetic looking FG procedural textures with strange effects.


Nah, usually it ain't - not in what I have seen from up-close. Actually on FSWeekend helicopter pilots used to FSX really loved the close-distance visuals FG could provide - they hadn't experienced that so far.
Thorsten
 
Posts: 12490
Joined: Mon Nov 02, 2009 9:33 am

Re: We'll soon need a MUCH better terrain engine :-)

Postby vnts » Sat Dec 11, 2021 11:19 am

V12 wrote in Sat Dec 11, 2021 9:37 am:1. LOD system - 16 years ago FSX had very effective LOD system [..]

Quote what/who you are responding to? I didn't mention P3D or FSX LoD in my last post.
V12 wrote in Sat Dec 11, 2021 9:37 am:2. P3D has incredibly expensive reflections - hmmm, why I did not observed incredibly fps drop when I activate reflections ?

Rendering the scene again is expensive for the benefit of the effect. Whether you experience drops depends on just how much everything else costs in P3D. It also depends on where your bottlenecks were with different (CPU/GPU) parts of P3D before and after reflections were activated.

Google P3D reflections and the first result is a thread titled "Dynamic Reflections - Extreme FPS impact (v4.5)" which advices the poster "For dynamic lighting, and for dynamic reflections, you'll need a high-performance computer. And that would mean a desktop, running on the sweet side of 4.5 GHz." and "It's been said here that P3D doesn't like Dynamic Lighting [..]".

If you follow the way rendering works, it's possible to tell this approach has problems without needing to test. But it's a convenient quick feature to for game companies to dump performance if other graphics features are plain and they don't want to develop - like enabling RTX on old games, or hoping PC people use 4k screens with last gen console games. There are other approaches, but that needs a company that wants to develop graphics tech, in which case P3D would look very different.
V12 wrote in Sat Dec 11, 2021 9:37 am:3. Art addons - I use only ORBX Global, no EnvTex, Tomato or something other

Did any addons improve the scenery on display? If so, my point was that P3D, a commercial product, looked worse than the image, and had to be improved. That doesn't say much for the quality of a product simply by being commercial.
V12 wrote in Sat Dec 11, 2021 9:37 am:4. Low texture resolution - hmmm, I do not observe it, for me, it is OK.

I can see it on your images even from far away..that's not a high standard for graphics.
V12 wrote in Sat Dec 11, 2021 9:37 am:P3D / FSX with ORBX have more than 9 GB base textures. How much textures are in FG ?

The point was that the commercial tech behind P3D is dated. FSX is extremely dated and hasn't been developed at all in a long, long time - just the odd maintenance. Adding more art (including paid art addons from 3rd parties) to brute force graphics doesn't change that.
V12 wrote in Sat Dec 11, 2021 9:37 am:5. "... on good developed areas" - P3D with ORBX looks reasonable good around the all world.

Yes - my point was that it's the art/content contributions that was historically lacking in FG, not the tech, despite FG not being commercial. Your "reasonable good" is not high for 2021 - have a look at some game engine rendering demos - even from many years ago (it's a bit hard to find ones that are in a simulator "real" style, as a lot of games use a painterly/cartoonish/stylised art style - for lots of design or business/profit reasons including simpler visual language to assist gameplay. Architecture or movie demos done in real-time game engines often have a more realistic art style.).
V12 wrote in Sat Dec 11, 2021 9:37 am:FG has 2 big advantages - grass overlay and smooth snow effect based on weather.

The situation with P3D is LM brought part of the old FS-X license, and management at some corner of the company is mainly just using the LM name and access to market it for simple training and whatever. It was 32 bit until a few years ago. There has been very little tech development. They've been content to get a cut of addons, and turn a blind eye to people using it for non-professional use (like your use of the Academic license). Addon makers have been slow to change to XP-11, with more recent tech, as FS-X/P3D has the monopoly, and they continue to get decent sales if they stick with the current system. So that keeps people using FS-X/P3D. X-P 11 forced a small change in P3D, and the quick attempt by a game studio to add a DX-11 openworld renderer to the FS-X codebase is showing how neglectful they've been. The majority of the profit LM made from P3D over the years would be funneled into the budgets of other non-software projects LM are doing (and in commercial companies a fraction goes to shareholders as dividends). It's just commercial motives at work.

Kind regards
vnts
 
Posts: 409
Joined: Thu Apr 02, 2015 1:29 am

Re: We'll soon need a MUCH better terrain engine :-)

Postby S&J » Sat Dec 11, 2021 6:18 pm

vnts wrote in Sat Dec 11, 2021 11:19 am:Yes - my point was that it's the art/content contributions that was historically lacking in FG, not the tech, despite FG not being commercial.


Shouldn't this be addressed ?

What is it about FG that brings in 'the tech' development whilst inhibiting content developers ?

Shouldn't there be parity ?
What are the longer term ramifications for FG if content development continues to be 'lacking' ?
"Stay away from negative people.They have a problem for every solution." - Albert Einstein
S&J
 
Posts: 794
Joined: Wed Aug 26, 2020 7:31 pm

Re: We'll soon need a MUCH better terrain engine :-)

Postby Thorsten » Sat Dec 11, 2021 7:24 pm

Shouldn't this be addressed ?


/rant on/

For years I've been posting, writing. saying - people, it's easy - get a good photograph of a few trees in your area, make a tree texture sheet - and we have accurate trees. I've written a detailed tutorial how to customize landcover, what to do and what not to do, posted on the newsletter, explained countless times that it's really easy.

The effect? Zilch. Nada. Niente. Zero.

So no - this should not be addressed FG-side, this has been addressed and addressing it has failed miserably.

It's really simple - if you don't like what you're seeing - bring better textures. Customize effects. Make a merge request. Under no circumstances, ask, suggest or even think that someone else should do this. Or fly in the areas where people who give a crap have invested their work.


What is it about FG that brings in 'the tech' development whilst inhibiting content developers ?


The mindset of 'I don't understand all this complicated stuff, I'd rather complain to someone and whine and be really annoying - maybe that'll get it done.'

What are the longer term ramifications for FG if content development continues to be 'lacking' ?


Again, very simple - you have to use what people who give a crap take care to create - if your favourite scenery/aircraft/... is not in there, you're just out of luck. End of story.

rant ends//

Seriously...
Thorsten
 
Posts: 12490
Joined: Mon Nov 02, 2009 9:33 am

Re: We'll soon need a MUCH better terrain engine :-)

Postby vnts » Sun Dec 12, 2021 3:19 pm

S&J wrote in Sat Dec 11, 2021 6:18 pm:
vnts wrote in Sat Dec 11, 2021 11:19 am:Yes - my point was that it's the art/content contributions that was historically lacking in FG, not the tech, despite FG not being commercial.

Shouldn't this be addressed ?
What is it about FG that brings in 'the tech' development whilst inhibiting content developers ?

In addition to what Thorsten wrote..

There are contributor issues on many scales that can affect contributions to any specific area of the project - for example in the largest scale: if all the social dynamics remained unchanged, but the awareness/profile of the project increased the amount of people finding and using about FG by 2x, there might be 2x more contributors.

Looking at the smaller scales - of contributions to regional materials by people that already follow FG closely i.e. power users or existing contributors to 3dmodels/textures/aircraft :

- The tech uses the approaches taken by photorealistic rendering for art/movies, and also by game engines with advanced graphics tech (here is a pre-historic 14 year old SIGGRAH paper from one of the big game engines roughly around the time FS-X released from google: link). For example, if a movie needed photorealistic terrain that didn't exist in the real world or wasn't convenient to film - e.g. a sci-fi or fantasy world - they would use software like Terragen or Vue (link, link) to create scenery using these types of procedural techniques. This approach has a really high quality ceiling. They'd never try to do what MS-FS lazily did - create a simple top-down image with only 8-bit RGB colours for surface material properties (like an orthophoto), and just apply it even on steep slopes/cliffs regardless of surface material. That approach hits a big limit on how realistic it can look. For example, there are no rendering papers I've come across about this approach .

- The existing tech massively multiplies contributions . It uses things people involved with 3d models/textures/aircraft already use. The workflow is basically get an idea of what scenery actually looks like: from local knowledge, from using instantstreetview's highest zoom which seems to be high quality aerial/done photos in some areas, from the streetview photos which have line of sight on relevant terrain, from google image search etc. And then tweak numbers up and down in an XML file until it looks right. Press reload materials, reload scenery. That's it. Existing textures can be used, recoloured/resynthesiszed , or new textures can be created. Recent advances in tools mean people can avoid learning to 'draw' with a mouse.

- Incoming people probably think of available contribution workflows in other commercial flight games with dated tech. They don't fully grasp that the terrain mesh, or triggering a scenery build refresh, isn't everything that contributes to visuals. Newcomers seem look first to generate terrain to improve visuals (I don't mean airports). And then when they find it doesn't hugely transform visuals - after they put in a lot more effort than tweaking a regional materials text file - they probably put it down to FG being WiP and limited. For the future, it /maybe/ doesn't help that MS-FS spent millions on advertising and talking about photoscenery - future incoming people may have decided FG can't match terrain quality given the lack of opensource photoscenery sources with the same resolution/coverage, or bandwidth costs being too high - MS-FS supposedly needs 2 petabytes streamed from servers (a huge marketing budget is typical for big game companies, where some games spend up-to 4x the development cost on marketing (link) - as it's easier to throw money at marketing, compared to managing development of better products).

- There was likely some element of 3d art/texture/aircraft contributors wanting to focus on their main projects, and not devote a fraction of the effort to the larger/shared FG project issues - thinking there was an existing team of 3d art/texture/aircraft contributors who will take care of their part of the world. A small fraction of effort would make FG first impressions better all over the world, which in time would bootstrap itself with increased contributors, and a fraction of them improving scenery. The situation is a bit like with the promotional parts of FG that Thorsten used to do before he was busy - articles on flightgear.org, release posts, articles flightsim.com and the forum there etc. If communicating information about the project to the world drops to zero, then it will - fairly invisibly - affect the project in the long run. This can happen if everyone assumes there's a team already doing it. Since Thorsten got busy it has dropped off - James and I did get at least the release post and I did a write up on Keflavik/Iceland - with others contributing screenshots and help. But aside from that, it dropped off somewhat.

- In addition to the above point, as Thorsten mentioned, thinking this area of development looks (very superficially) complicated , when the controls exposed to scenery developers to tweak are actually simpler than aircraft XML files for a lot of things - just tuning numbers up and down. I've thought that with the planned in-sim GUI , a set of sliders , and a way to activate tweaking of a specific landclass could help - but that only improves some parts.

Kind regards
vnts
 
Posts: 409
Joined: Thu Apr 02, 2015 1:29 am

Re: We'll soon need a MUCH better terrain engine :-)

Postby merspieler » Sun Dec 12, 2021 3:32 pm

vnts wrote in Sun Dec 12, 2021 3:19 pm: The situation is a bit like with the promotional parts of FG that Thorsten used to do before he was busy - articles on flightgear.org, release posts, articles flightsim.com and the forum there etc. If communicating information about the project to the world drops to zero, then it will - fairly invisibly - affect the project in the long run. This can happen if everyone assumes there's a team already doing it. Since Thorsten got busy it has dropped off - James and I did get at least the release post and I did a write up on Keflavik/Iceland - with others contributing screenshots and help. But aside from that, it dropped off somewhat.


What about the monthy newsletter on the wiki? That could be cross-posted to the forum and/or flightgear.org
Nia (you&, she/her)

Please use gender neutral terms when referring to a group of people!

Be the change you wish to see in the world, be an ally to all!

Join the official matrix space
merspieler
 
Posts: 2243
Joined: Thu Oct 26, 2017 11:43 am
Location: Wish to be in YBCS
Pronouns: you&, she/her
Callsign: you&, she/her
IRC name: merspieler
Version: next
OS: NixOS

Re: We'll soon need a MUCH better terrain engine :-)

Postby Thorsten » Sun Dec 12, 2021 4:28 pm

What about the monthy newsletter on the wiki? That could be cross-posted to the forum and/or flightgear.org


A promotional piece is rather different from a collection of announcements and screenshots. It needs to be thought through, it needs to be an interesting read, tell a story while at the same time provide information. Ideally it leaves people with the impression of having learned something interesting and wanting to try it out.

If you look e.g. at the Space Shuttle Piece on the FG page, you'll see that this contains original quotes from Astronauts I've looked up as well as a screenshot collection that has taken a month to assemble.

You can cross-post all sorts of thrown-together bits and pieces - but it's not going to make something equal to a promotional piece. In fact, it might look rather unprofessional and put people off.
Thorsten
 
Posts: 12490
Joined: Mon Nov 02, 2009 9:33 am

Re: We'll soon need a MUCH better terrain engine :-)

Postby wkitty42 » Sun Dec 12, 2021 7:04 pm

Thorsten wrote in Sun Dec 12, 2021 4:28 pm:If you look e.g. at the Space Shuttle Piece on the FG page

wow! i didn't realize that was 6 years old, now...
"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: 9148
Joined: Fri Feb 20, 2015 4:46 pm
Location: central NC, USA
Callsign: wk42
Version: git next
OS: Kubuntu 20.04

Re: We'll soon need a MUCH better terrain engine :-)

Postby vnts » Mon Dec 13, 2021 9:35 am

merspieler wrote in Sun Dec 12, 2021 3:32 pm:
vnts wrote in Sun Dec 12, 2021 3:19 pm: The situation is a bit like with the promotional parts of FG that Thorsten used to do before he was busy - articles on flightgear.org, release posts, articles flightsim.com and the forum there etc. If communicating information about the project to the world drops to zero, then it will - fairly invisibly - affect the project in the long run. This can happen if everyone assumes there's a team already doing it. Since Thorsten got busy it has dropped off - James and I did get at least the release post and I did a write up on Keflavik/Iceland - with others contributing screenshots and help. But aside from that, it dropped off somewhat.


What about the monthy newsletter on the wiki? That could be cross-posted to the forum and/or flightgear.org

As Thorsten said, things aimed at people who don't know about FG, or people in the media, are a lot less casual than the news;letter - which can contain the minimum needed to point FG users towards other material/projects.

However, a link to the wiki newsletter could be added prominently in the forum I guess - so it would inform people that missed developments, or didn't know about the newsletter .

I meant more that articles on flightsim.com were missing compared to previously (despite the editor reaching out and asking, so there is interest..) .Thorsten also posted on the flightsim.com forum - and the flightsim.com people made a FG subforum there.

The flightsim.com forum is probably also an example the phenomenon of people focusing entirely on their main projects compared to devoting a tiny fraction shared FG stuff - where all that's needed is a tiny fraction of time if a sizeable percentage of contributors did it - Thorsten made a thread about posting there so people could let the large community there know about some of the big projects they are working on (FG does a lot of unique, or high quality things, and word would spread) : viewtopic.php?f=42&t=34444

I think the newsletter is done a bit quickly/casually for flightgear.org for people who aren't already using FG. But another possibility is to embed the monthly newsletter in the launcher. This could be done by making a locked down and vandal-proof version of the newsletter in the wiki. Then embedding a browser page in the qt-launcher - or write the newletter using a wiki markdown plugin that, is also trivially render-able in the launcher via a qt plugin.

I've mostly just quickly written about rendering things I'm following closely to the newsletter - and just posted any development videos I noticed in the youtube results for the last month. Usually I end up adding things just before the end of the month, or when I have time/remember in the first days of the month :mrgreen: . There are a couple of contributors adding to the newsletter of late. To be worth getting into the launcher, it might need a bit more people posting about other areas/projects they're involved in, or following closely. But the huge exposure and likely a large response to postings would motivate people. The QGIS application, of all things, turned out to be an excellent example of in-application community news, updates about events, patches etc.

----

And come to think of it the newsletter is also another example of this phenomenon actually - there was a thread about it some years ago on the project subforum talking about the social dynamics: viewtopic.php?f=42&t=32405

But those days editing the wiki was far more difficult. The new WYSIWYG wiki is easier than the forum - there's even drag and drop functionality for uploading screenshots but you need to drop it directly onto the blue button. So there have been more contributions, including from newcomers .
Skimming the thread:
Thorsten wrote in Sat Jun 24, 2017 10:56 am:After some reflection:
To a first approximation, it's fairly reasonable to assume that (some) other people tick pretty much like you do. So if you don't want to do some task for some reason, chances are others do not either. If you're waiting for someone else to do something, chances are that others are also waiting.

The trick is to recognize we don't get anything done by waiting for someone else to do it.

So the solution is that if you think something is important/useful etc., you go ahead and do it if you can spare the time (and before you go claim that I preach water and drink wine, I spent the last days documenting how to add effects to an aircraft as well as bringing the Vostok wiki up to speed).

It's a wiki - it makes no sense to suggest e.g. to copy devel notes to the changelog - you can go ahead and do it if you believe it ought to be done - don't suggest that someone else does it, this makes no sense in a wiki - remember all others are waiting for you to do it - because they by and large tick the same way.

wlbragg wrote in Sun Jun 25, 2017 8:08 pm:I see yet other "enthusiasts" that do both and also step up to do things solely because they know that it helps the project not because they want to. I strive to emulate them.

These people are the glue that is keeping this project going. I think of where FG would be if these individuals had never contributed anything or decided to no longer contribute.

It's an ongoing issue in the FG project, "willing human resources".

Hooray also pointed out that some shared work, or "glue" contributions, for newsletter type things, was making it clear to people that the work they were doing were doing qualifies for the newsletter, and writeups are fine for non-english speakers - and probably people feel a bit shy about adding their own projects to a newsletter or promotional articles, or posting a thread about it in the flightsim.com forums:
Hooray wrote in Sat Jun 24, 2017 3:26 pm:...
In fact, I ended up using a simple template that basically invited people to contribute to the newsletter, pointing out how to get started doing this, which included links to basic introductory articles, including instructions on adding screenshots/images and youtube video. I also made it clear that the newsletter was community-driven and that everybody was invited to contribute to it, also people who stated that their grasp of the English language would not suffice, because we had others (like e.g. Red Leader) who were willing to help review/proof-read such contributions.

These days however, I am not as much involved in wiki/FG matters, and I guess that nobody is reaching out to our contributors anymore to get involved in the newsletter ?
...
Then again, I am certainly not opposed to any efforts surrounding improvements to the newsletter, I just don't believe that it's necessary, or that it's going to work out too well - like I said, back when the newsletter was apparently working "well enough", there was quite a bit of behind-the-scenes networking going on, including 1:1 mentoring to get new contributors up to scratch with the wiki.

Finally, let me state that I would love to see the newsletter alive and kicking again, mainly because I am also not very involved in FG matters these days, and I find it annoying to have to browse the archives (forum/list) or the commit logs to come up to speed with FG matters.


Kind regards
vnts
 
Posts: 409
Joined: Thu Apr 02, 2015 1:29 am

Re: We'll soon need a MUCH better terrain engine :-)

Postby merspieler » Mon Dec 13, 2021 9:53 am

Thanks for explaining.

Adding the newsletter to the launcher (or just even linking it) could get users attention to the help wanted section.
Nia (you&, she/her)

Please use gender neutral terms when referring to a group of people!

Be the change you wish to see in the world, be an ally to all!

Join the official matrix space
merspieler
 
Posts: 2243
Joined: Thu Oct 26, 2017 11:43 am
Location: Wish to be in YBCS
Pronouns: you&, she/her
Callsign: you&, she/her
IRC name: merspieler
Version: next
OS: NixOS

PreviousNext

Return to Hangar talk

Who is online

Users browsing this forum: No registered users and 2 guests