Board index Other Hangar talk

Help needed - market research for FG

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: Help needed - market research for FG

Postby Bjoern » Tue Nov 20, 2012 1:19 pm

dbelcham wrote in Mon Nov 19, 2012 9:45 pm:2) [...]


I have to agree with that.
When scripting the installer scripts, I just did not know which TerraGear repo was the "right" one to use.

Condensing information and use of repositories might just be the way to go there as well.
Do a barrel roll!
User avatar
Bjoern
 
Posts: 384
Joined: Fri Jan 06, 2012 10:00 pm
Location: Nearest: THF
Version: Next
OS: ArchLinux, Win 7

Re: Help needed - market research for FG

Postby dbelcham » Tue Nov 20, 2012 1:57 pm

Thorsten wrote in Tue Nov 20, 2012 7:50 am:So editing scenery by hand is something that is fairly irrelevant in the overall scheme of things, and hence not well supported.


Is it possible to edit scenery by hand? Yes. Is it desirable to edit scenery by hand? I'd say yes. Fly over the central US or any of Canada and tell me that there isn't a need to edit those areas by hand. There has to be a reason that people are hand editing 1 degree blocks and I'd venture to say it's because the base scenery (airports and terrain) is not adequate. So regardless of what the giant batch process to generate new scenery does the lack of detailed landcover and detailed DEMs is a negative for FG. So if there already is a movement to improve those problems one tile at a time, why shouldn't it be embraced and made as friction free as possible?

What is wrong with having a FG equivalent of this (http://developer.x-plane.com/tools/worldeditor/) X-Plane page? For the basic shell all that is needed is something on the wiki. The harder part is getting the release packages for all of the tools. For that the teams/people working on those tools have to start thinking about releasing software instead of just building it. The main package for FG does this. You can go and download the official FG installer from either the website or from Jenkins. If it can be done for that, which arguably is the most complex release package of any project in FG, they why can't it be done for the other tools? The only reason I can think of is that the different projects have a developer-centric view.

As soon as you start mentioning things like
Bjoern wrote:I just did not know which TerraGear repo was the "right" one to use.

you should be able to tell that things are too developer-centric. Any mention of Version Control Systems in the process of installing and running an application will lose 90%+ of all people who want to attempt to do these things. And since the efforts that they would make building better scenery is something that would benefit us all, then by blocking the entry of those 90% we are essentially saying we don't want to see improvements to FG scenery/add-ons.

I'm not suggesting that everyone's world needs to be turned ass over tea kettle. For some groups it will, but one can argue that the cowboy style development (all code changes being committed in the trunk for example) needs to be reined in and made more professional if there is any desire to grow the content creation community and, ultimately, the content in FG. All I'm saying is that there should be an official place to download releases of developer tools that doesn't require Version Control Systems, cmake and other developer gymnastics. If the tooling is something that may be used by the community to improve FG then it should have easily downloadable and runable official releases.
dbelcham
 
Posts: 58
Joined: Thu Jun 07, 2012 3:55 pm

Re: Help needed - market research for FG

Postby Thorsten » Tue Nov 20, 2012 6:33 pm

Is it possible to edit scenery by hand? Yes.


Not on a meaningful scale.

Is it desirable to edit scenery by hand? I'd say yes. Fly over the central US or any of Canada and tell me that there isn't a need to edit those areas by hand.


There isn't any need to edit central US or Canada by hand. As Pacific Northwest scenery or Colorado proves, using higher resolution geodata does the job just fine. To say this bluntly and even a bit provocative - every developer doing scenery tools would be wasting his time to create a manual scenery editor - that same time spent to improve the geodata toolchain processing will give you a factor 10.000 more bang for the buck.

My understanding is that world scenery is bad because it is old, because back then graphic cards could not handle a million vertices, so the mesh was intentionally not created with full possible resolution, and it is bad where there's no GPL licensed data. CORINE data can create compelling scenery without hand editing - custom Italy has at least the quality of Innsbruck, if not better.

If you want to edit the vast areas of central US by hand, good luck - an Innsbruck sized chunk doesn't make much of an impact.

What is wrong with having a FG equivalent of this (http://developer.x-plane.com/tools/worldeditor/) X-Plane page?


It costs time to create, which is better spent elsewhere. If you can get orders of magnitude more output by geodata processing, this is what developers should focus on.

So if there already is a movement to improve those problems one tile at a time, why shouldn't it be embraced and made as friction free as possible?


Because even if you have 1000 people in this movement (which you haven't), improving the geodata approach is still more efficient.
Thorsten
 
Posts: 11086
Joined: Mon Nov 02, 2009 8:33 am

Re: Help needed - market research for FG

Postby Hooray » Tue Nov 20, 2012 7:52 pm

improving tool chains for manual scenery editing will not scale as well as improving tool chains for automated scenery compilation. The latter only requires CPU power and time to run. While having a great scenery editor in itself will not give us anything, not without having a huge number of users willing to learn/use it, and we'd still depend on end users here.

It's basically like manual vs. procedural texturing: coming up with an algorithm to do something procedurally scales better than manually creating all the required textures, at some point there's no manual work involved any more "only" CPU/GPU computations.

If you don't believe it, just look at the AI traffic system: it's well documented, there are a bunch of tools to create AI flight plans manually - yet, the only scalable option to populate our AI world is using code to do this procedurally, which work "well enough" for most cases - instead of being "perfect" for just a single case.

There are actually countless examples to be found here... just look at FlightGear's release history: Only after the build server was added, did we get to see regular FG releases and a well-defined release schedule/plan. Previously, there were no regular releases, and creating new releases always meants tons of manual work for certain people.

So the key thing really is LARGE-SCALE AUTOMATION, in all areas possible ...
Please don't send support requests by PM, instead post your questions on the forum so that all users can contribute and benefit
Thanks & all the best,
Hooray
Help write next month's newsletter !
pui2canvas | MapStructure | Canvas Development | Programming resources
Hooray
 
Posts: 11340
Joined: Tue Mar 25, 2008 8:40 am

Re: Help needed - market research for FG

Postby dbelcham » Tue Nov 20, 2012 11:13 pm

As my last post on the topic imagine this situation (which is the current situation).
- The landcover for Canada is not very detailed. For example it is missing an entire 25,000 person city at the end of one of the runways at CYEG. The level of detail is for each landclass is aweful. Entire 1000 acre forested areas are missing. Lakes that are multiple square km near CYEG are missing.
- There is extremely detailed landcover data available for the entire country of Canada. There's a definite chance that while the data is free for use it won't meet the licensing requirements to be included in the main scenery repository. The shapes aren't perfect but in less than 1hr I was able to clean up a 1x2 degree area to the point where it is a hundred times better than the default scenery.
- There's higher resolution DEM data available. Just like the landcover, the licensing is free for use but might not be compatible.
- There's public domain information available for every radio/cell tower in the country that is over something like 10m in height. You can find the exact lat/lon as well as the exact height.

With some work to clean up the landcover (which is a completely achievable goal) there's no reason that this information couldn't be run through the tooling using a simple script (which I've written and multi-threaded where appropriate). I can then release the scenery under an appropriate OSS license so as not to conflict with the original data licenses. The community wins when this happens does it not?

Based on the opinion of you two, there's no reason that the community should support this type of effort. I understand that now. I won't pursue such an effort any further. Consider the project ended or at the least the results internal and private.
dbelcham
 
Posts: 58
Joined: Thu Jun 07, 2012 3:55 pm

Re: Help needed - market research for FG

Postby Thorsten » Wed Nov 21, 2012 7:48 am

Based on the opinion of you two, there's no reason that the community should support this type of effort.


That's not manual scenery creation you're talking about, that's geodata processing. I think I gave a very clear opinion to the contrary of how you quote me, i.e. that the community should support geodata processing rather than manual scenery creation.

The community wins when this happens does it not?


Opinions are divided on that. Some folks (myself included) are in favour of custom scenery projects even if they're not GPL because they enhance the attractivity of Flightgear. The scenery devel team is of the opposite opinion that all efforts should be focused to create a GPL compatible hires world scenery, that support for custom scenery takes much-needed devel time from the actual goal and hence does usually not support custom scenery projects. So one might argue that in their view the community wins more in the long run without custom scenery creation. (Please don't argue this with me, I'm just spelling out the positions as they've been exchanged on the mailing list - neither can I do anything about it, nor am I going to defend someone else's opinion here).
Thorsten
 
Posts: 11086
Joined: Mon Nov 02, 2009 8:33 am

Re: Help needed - market research for FG

Postby Hooray » Wed Nov 21, 2012 8:23 am

finding myself in agreement with Thorsten here: I really DO appreciate the manual scenery editing efforts, like for example LOWI *very much*.
However, the truth is: it simply doesn't scale too well. So we need to find less specific solutions that are more generic, but still provide an acceptable result, with little manual work required - preferably, only the work required to code the tool chain, and nothing else.
Please don't send support requests by PM, instead post your questions on the forum so that all users can contribute and benefit
Thanks & all the best,
Hooray
Help write next month's newsletter !
pui2canvas | MapStructure | Canvas Development | Programming resources
Hooray
 
Posts: 11340
Joined: Tue Mar 25, 2008 8:40 am

Re: Help needed - market research for FG

Postby Gijs » Wed Nov 21, 2012 10:00 am

dbelcham wrote:There's public domain information available for every radio/cell tower in the country that is over something like 10m in height. You can find the exact lat/lon as well as the exact height.

Wondering how this fits the landcover discussion... :wink:

Anyway, could you please share this source, or submit the objects yourself? Adding shared objects (like windturbines, power pylons and antennas) is easy, especially now that we have a nices forms at http://scenemodels.flightgear.org/contribute.php
Do note that we already have FAA obstruction data included, so most of the (taller) antennas should be present already (see this list).
Airports: EHAM, EHLE, KSFO
Aircraft: 747-400
User avatar
Gijs
Moderator
 
Posts: 9380
Joined: Tue Jul 03, 2007 2:55 pm
Location: Amsterdam/Delft, the Netherlands
Callsign: PH-GYS
Version: Git
OS: Windows 10

Re: Help needed - market research for FG

Postby Bjoern » Wed Nov 21, 2012 7:34 pm

Hooray wrote in Tue Nov 20, 2012 7:52 pm:So the key thing really is LARGE-SCALE AUTOMATION, in all areas possible ...


Have I said that I love this approach about FG?
Automate EVERYTHING!



dbelcham wrote in Tue Nov 20, 2012 11:13 pm:Based on the opinion of you two, there's no reason that the community should support this type of effort. I understand that now. I won't pursue such an effort any further. Consider the project ended or at the least the results internal and private.


Oh come on!
Just make it available somewhere, link to it in a wiki article or open a forum thread about it and wait what happens.

You've invested at least a few hours into this, so why not share it with those interested?
Do a barrel roll!
User avatar
Bjoern
 
Posts: 384
Joined: Fri Jan 06, 2012 10:00 pm
Location: Nearest: THF
Version: Next
OS: ArchLinux, Win 7

Re: Help needed - market research for FG

Postby Hooray » Wed Nov 21, 2012 9:04 pm

Regarding the level of detail, accuracy, fidelity - manual scenery editing efforts will always be superior to automated large-scale scenery creation.

However, we usually only have a handful of people willing/able to dedicate the time required to do this.

The results are definitely appreciated, however we need to have a solution for the 99.995% of areas for which we don't have dedicated teams of scenery developers.
Just look at great examples like LOWI: it took tuxklok and ot-666 several years to complete this, and they are still working on it today.

So instead of developing aircraft, they develop scenery - but they are unusually dedicated. We don't have many other scenery developers contributors who are fully dedicated to a single thing/place/aircraft.
Please don't send support requests by PM, instead post your questions on the forum so that all users can contribute and benefit
Thanks & all the best,
Hooray
Help write next month's newsletter !
pui2canvas | MapStructure | Canvas Development | Programming resources
Hooray
 
Posts: 11340
Joined: Tue Mar 25, 2008 8:40 am

Re: Help needed - market research for FG

Postby gluon » Mon Dec 10, 2012 2:13 pm

To revive this thread, here are my 2¢:

+1 for Thorsten's and Hooray's points.

Price isn't really a selling point to hard-core FSX users. FSX users are heavily invested in the platform. Some of them have spent alot of money on add-ons. So the hardcore users are unlikely to change even if it's free (cognitive dissonance).

Additionally, many of the positive points of FlightGear presented in the past by the FG community have a technical focus. This has already been mentioned here, too. Something I've often read in other sim forums is (Quote) "FlightGear is a flight simulator for people who like to develop flight simulators". The point is not completely wrong and demonstrates that there is a lack of end-user sensitivity within the community. Which also has been discussed here.

Which brings me to the next point: User experience.
Something I've repeatedly read in forums is how difficult it is for new users to get started. For example, after starting FG most of the default aircrafts (including the C172) is sitting on the runway with motors off. Besides not being a very realistic situation, a new user is confused in this situation. In X-Plane and FSX planes are ready to fly at default start-up.
Additionally, all aircraft commands are not accessible via the UI - you need keyboard-shortcuts which the user doesn't know at this point. To add to the confusion, the menubar is automatically hidden (how can the user find the necessary information?). A "First-start" dialog, which is common in many applications today, could remedy this situation (the info message appearing at start-up doesn't really help - which invisible menu should the user click?). Many users also complain about missing or poor documentation and translations.
In this regard FlightGear is like a hidden gem: There are many cool features which unfortunately only a few people know about.

Or take the F14 which has empty fuel tanks upon the first start. No wonder people complain it doesn't work.
Or the default C172 which banks so heavily to the left - the X-Plane or FSX C172 don't do that. Even if it's realistic: Is a first-time user really going to expect this to be realistic or quite simply a bug?

If users manage to make his/her way through this situation, what do they experience? The default scenery has been discussed. Let me mention, that FSX has a sort of default scheme for building models of airports, which has been mentioned in some magazine articles comparing FG to FSX. So there are no empty airports, which is neat.

In addition, despite having a high quality texture option delivered with FG (the DDS textures), the user is presented with the default texture-set which is of low resolution and not well matched. Thus, the visual appearance (eye candy) is worse than FSX or X-Plane by default. However, with the DDS textures it is superior and almost at FSX Add-On level. This way FlightGear is sold below its value. So why not use them as default? Or at least use a scaled down version of them as the default, if memory usage is considered to be the road block.

However, and this also has been already mentioned, apparently performance is not a focus of the development anyway. It runs with 20-30 frames on my fairly recent machine. Add Rembrandt to the equation and I get a dia-show. This also goes along the line Björn says: Be honest about what you sell. You'll hear the statement often in the FG community that it will run on old hardware. Yes, it may, but the default set-up needs a bit more power and the users really don't know how to tweak the setup easily. In fact, FSX runs more smoovely here (and it is known to be a performance hog...).

FSX in contrast also has easily customize settings: e.g for graphics, scenery, traffic, weather. No need to edit text files. The settings can be adjusted using a GUI featuring two levels: Overview and detailed. The overview level allows for setting general performance of a group of settings (e.g. graphics). The detailed for setting individual detail setting.


The situation if a user persists and goes looking for add-ons has also been mentioned: Quality add-ons, like aircraft or custom scenery) is hard to find. Unfortunately many of the good aircraft are kept in separate hangar-sites. If they are/will be included in the official repository is unknown. Note that this situation is very similar to the experience of FSX users.

A way to improve the aircraft download site would be to filter planes by rating (which has been already done by Gijs I think). In addition, I would suggest to sort the aircraft into categories: e.g. general aviation, airliners, military, historic, warbirds, helicopters. Often I am looking for a plane of a certain type. And scrolling through the very long alphabetic list sure is not very user-friendly.


Here's some other random features I like in FSX that are not implemented in FlightGear:

* A flight analysis featuring a map displaying the flightpath and a vertical analysis graph and VCR-like playback.

* ATC: FS9 and FSX have a automatic ATC system where you can communicate with the towers and the regional ATC. In addition to the text it also sports different voices for pilots and towers so you can actually hear the radio communication. The voices are pre-recorded and sound natural (not robotic TTS). While the voices are nice, a text-based version would be sufficient. In FlightGear comparable ATC support has been broken since 2.0 and has only worked at approach.

* Multiple views: FSX has the ability to open multiple independent sub-views. So for example you can have the cockpit in the main window, a chase-view and a fly-by view in two sub-views embedded in the main view. This is a really nice feature I use regularly.
In FlightGear of course something comparable is possible if you hack and define multiple-views/cameras. However, its not accessible to everyday users (non-developers). In addition, all of the views are part of a camera-group and thus depend on each other. If I change the direction in one view all other views are also affected. If you switch to a different main view everything goes wild. So it only is suitable for static setups(~> Lelystad). Furthermore, you cannot reuse the standard views, such as fly-by view, chase-view or any of the additional views defined by the aircraft-model.

* Sounds: Sounds in FSX are in general much better. It features high-quality sound samples which don't sound repetitive. This adds to the experience and enjoyment. Also, the Doppler-shift employed to FG for outside views has a strange quality. I think there's some phase shifting going on there. Maybe its physically correct, but it sounds wrong to me. FSX does also feature Doppler shift which sounds natural.

* Eye-candy: The water-shader in FSX actually shows reflections of clouds and scenery. Though not the best (X-Plane is much better), this really adds to the experience. The streets are populated with cars that are moving. Landclasses are mostly seemless. Nice looking shorelines and animated waves incoming to shore.


On the positive side, the Helicopters are much more realistic in FG than in FSX. Also, with custom scenery and good textures it looks much better than FSX. I just hate those mushy FSX textures. The clouds and the weather simulation in FlightGear are far beyond what add-ons provide for FSX. As are the sunrises. And the atmosphere shader.
If you change weather settings, the transition between scenarios is smooth and not abrupt.

In fact FlightGear is like a diamond in the rough - with a little bit of polish it could be amazing.

Overall, this post is so negative, so don't get me wrong. I love FlightGear, so much I have been around the simulator for 12 years! In fact it was my first flight sim. I am also very exited about the quality and the recent developments of FlightGear. And I'd like other people to share this excitement.
gluon
 
Posts: 68
Joined: Tue Aug 18, 2009 7:43 am

Re: Help needed - market research for FG

Postby Gijs » Mon Dec 10, 2012 2:44 pm

Agreeing on most (or all) points made, but there is one thing I'd like to correct:
To add to the confusion, the menubar is automatically hidden (how can the user find the necessary information?).

Autohide of the menubar is disabled by default. However, if you do enable it (accidentally?), it gets enabled on the next sesion as well (settings are stored between sessions). That's with a lot of things actually (rendering settings, certain aircraft's fuel weights etc.).
Airports: EHAM, EHLE, KSFO
Aircraft: 747-400
User avatar
Gijs
Moderator
 
Posts: 9380
Joined: Tue Jul 03, 2007 2:55 pm
Location: Amsterdam/Delft, the Netherlands
Callsign: PH-GYS
Version: Git
OS: Windows 10

Re: Help needed - market research for FG

Postby Hooray » Mon Dec 10, 2012 2:53 pm

I mostly agree with what you said. However, regarding DDS textures and other "high-end eye candy" features it's worth keeping in mind, that not all new FG users will have access to the latest hardware, in the form of GPUs like an NVIDIA680 with 3 GB of dedicated DDR memory. So we need to make compromises inevitably. That said, even today (with the last 2.8 release) there are many users left with a FG release that doesn't work "out of the box" for them, because of the stringent hardware requirements that are enabled by default.

So what's really needed is a form of dynamic feature scaling that can check the features supported by a platform and then find a working set of features which still provide reasonable performance and a good experience, all of this taking place during start-up.

gluon wrote in Mon Dec 10, 2012 2:13 pm:Which brings me to the next point: User experience.
Something I've repeatedly read in forums is how difficult it is for new users to get started. For example, after starting FG most of the default aircrafts (including the C172) is sitting on the runway with motors off. Besides not being a very realistic situation, a new user is confused in this situation. In X-Plane and FSX planes are ready to fly at default start-up.

I think we could take Stuart's new "Aircraft Checklists" (see the wiki) feature and extend it such that the checklists are purely procedural, so that they could not just be used as interactive checklists for humans, but also as checklists ("scripts") for the program (fgfs) to change the aircraft configuration accordingly - programmatically. So that "autostart" and similar features would no longer be purely based on custom Nasal scripts, but instead automatically be supported by all aircraft that provide a corresponding checklist, that could then be used by the simulator to switch to a certain state directly.


Additionally, all aircraft commands are not accessible via the UI - you need keyboard-shortcuts which the user doesn't know at this point.


This is something that could probably be solved through a Nasal layer that automatically provides UI bindings for all important hot spots.
In fact, one could even create dialogs with buttons procedurally for such key bindings. We have lots of Nasal/GUI examples doing this sort of thing. It would only be a matter of writing such a layer, that would then provide a simple UI for important bindings.

To add to the confusion, the menubar is automatically hidden (how can the user find the necessary information?). A "First-start" dialog, which is common in many applications today, could remedy this situation (the info message appearing at start-up doesn't really help - which invisible menu should the user click?).


That would be a very simple Nasal/GUI dialog to implement that would only need to use a "userarchive" attribute to determine if it's intended to be shown or not during startup.


Or take the F14 which has empty fuel tanks upon the first start. No wonder people complain it doesn't work.


In other simulators, there's a dedicated W&B dialog for fueling etc - this could also be scripted in a portable fashion using Nasal, but obviously we have different FDMs in FG, and we support very different types of aircraft. Providing a truly generic dialog would be difficult (at best) and would still require people to follow certain protocols. It would probably be better to implement a Nasal-driven skeleton that can be extended/customized by aircraft developers. Important concepts like "tanks" would then ideally be encapsulated already and provided with wrappers.

Or the default C172 which banks so heavily to the left - the X-Plane or FSX C172 don't do that. Even if it's realistic: Is a first-time user really going to expect this to be realistic or quite simply a bug?

Other simulators deal with this through the means of a "realism" slider/setting. I think we used to have a first stab at that in FG, but it never got truly implemented, right ?

If users manage to make his/her way through this situation, what do they experience? The default scenery has been discussed. Let me mention, that FSX has a sort of default scheme for building models of airports, which has been mentioned in some magazine articles comparing FG to FSX. So there are no empty airports, which is neat.


The problem here is our static scenery compilation process that depends on TerraGear - some other sims are able to create reasonable airports/buildings procedurally. In FG, that would be a huge paradigm shift.
A while ago, Thorsten and I suggested actually in a different thread that a fully scriptable random buildings subsystem could be used for many other neat things.
Procedural airport creation would just be yet another example here.

However, and this also has been already mentioned, apparently performance is not a focus of the development anyway.

Just to clarify: FG development generally isn't very focused at all - so it's wrong to suggest that "performance is not a focus" ;-)

It runs with 20-30 frames on my fairly recent machine.

That's clearly not very much, I am getting in excess of 40-55 fps and 20-40 ms on moderately recent hardware (without rembrandt!).
But I am also fairly familiar with tuning FG according to my needs ...

You'll hear the statement often in the FG community that it will run on old hardware. Yes, it may, but the default set-up needs a bit more power and the users really don't know how to tweak the setup easily. In fact, FSX runs more smoovely here (and it is known to be a performance hog...).

that's a good point, which I have also found to be true with X-Plane, too - FlightGear's defaults don't scale very well at the moment. But to be honest, backward compatibility has never been a real priority ...

FSX in contrast also has easily customize settings: e.g for graphics, scenery, traffic, weather. No need to edit text files. The settings can be adjusted using a GUI featuring two levels: Overview and detailed. The overview level allows for setting general performance of a group of settings (e.g. graphics). The detailed for setting individual detail setting.


I mentioned this earlier, I think that most rendering-related settings have been meanwhile been exposed to the GUI and made accessible there. The issue with the remaining settings is that they are simply not mutable at runtime, so that they require a full simulator restart - which is why only various launchers allow them to be changed. That's an architectural issue in FG, which isn't easily solved, but some folks are working on better runtime-reinitialization support. And in combination with the canvas system, it should definitely become possible to provide a neat UI then - for the moment, the GUI would be simple to do, but it could never work unfortunately ... it's similar with the frequently requested "save/load flight" feature: the GUI for this is the simplest part actually, but the architecture needs serious changes to support such features.


* A flight analysis featuring a map displaying the flightpath and a vertical analysis graph and VCR-like playback.

That should be very simple to implement using a combination of the canvas system and by accessing the built-in flight recorder subsystem.
Graphs can already be created using the canvas, we would just need a way to access the replay buffers from Nasal space.

* Multiple views: FSX has the ability to open multiple independent sub-views. So for example you can have the cockpit in the main window, a chase-view and a fly-by view in two sub-views embedded in the main view. This is a really nice feature I use regularly.
In FlightGear of course something comparable is possible if you hack and define multiple-views/cameras. However, its not accessible to everyday users (non-developers). In addition, all of the views are part of a camera-group and thus depend on each other. If I change the direction in one view all other views are also affected. If you switch to a different main view everything goes wild. So it only is suitable for static setups(~> Lelystad). Furthermore, you cannot reuse the standard views, such as fly-by view, chase-view or any of the additional views defined by the aircraft-model.

That's another long-standing feature request and which all core developers agree on actually. The thing is, it also requires some architectural changes: http://wiki.flightgear.org/Howto:Use_a_ ... siteViewer



I really don't think your posting is negative at all - it's definitely helpful to share such feedback, to ensure that FG becomes more attractive for people who are not "developers". So thanks for taking the time write all that! :D

The problem is the sheer amount of feedback here (4+ pages and counting), we need to start evaluating all the info and move it to the wiki or even the issue tracker to process it further.
Please don't send support requests by PM, instead post your questions on the forum so that all users can contribute and benefit
Thanks & all the best,
Hooray
Help write next month's newsletter !
pui2canvas | MapStructure | Canvas Development | Programming resources
Hooray
 
Posts: 11340
Joined: Tue Mar 25, 2008 8:40 am

Re: Help needed - market research for FG

Postby Philosopher » Mon Dec 10, 2012 3:59 pm

Gluon: I'm not quite getting the Cessna thing? Why would we have it be started automatically on start up? A) pilots in real life should know how to start their engines (since they definitely aren't started for them!) B) it is mentioned explicitly in the manual on how to start most piston engines ('}}}' then 's'), and the C182 is no different C) I'm sure there's a switch on the panel that says "Off L R Both Start" which should be an indication to click on it until it says "start" and D) there's also information on key bindings in Help->Common Aircraft Keys, so the user can just look there if they know that they need the magnetoes on. There are a couple aircraft that require other switches to be on (Piper Commanche, storch, Beaver, etc.), however, each of those aircraft document it very well under the aircrt help (key '?') so it shouldn't be an issue. Point is this: it's the user's fault if they can't start a piston engine! For jets, they either come already started or have autostart, and helis just need a key press of the '}' button. So I don't see why one would want the aircraft to be started when the sim is started up, other than a small matter of convenience :?:

Second Gijs' point about the menu: that should not be the case. As to some of your other points, a fair amount of those are being worked on, so it should improve over the next couple months, precisely to your point of a lot of recent development. I too am excited about flightgear's improvemts! Also, perhaps we should hold back on discussing this too vigorously until after the next release? Since it might fix some issues that have been mentioned in this thread.
Thanks,
Philosopher
(inactive but lurking occasionally...)
Philosopher
 
Posts: 1590
Joined: Sun Aug 12, 2012 6:29 pm
Location: Stuck in my head...
Callsign: AFTI
Version: Git
OS: Mac OS X 10.7.5

Re: Help needed - market research for FG

Postby Thorsten » Mon Dec 10, 2012 7:09 pm

In addition, despite having a high quality texture option delivered with FG (the DDS textures), the user is presented with the default texture-set which is of low resolution and not well matched.


Opinions are divided on the issue of the well-matchedness of the DDS texture set :-) Also, the higher resolution comes at the price of increased tiling problems (the texture resolution isn't exclusively in the file size, it's mainly the physical size onto which the texture sheet is mapped - and if you map to 500 m, you get a different result than if you map onto 2000 m). However, leaving that aside, the DDS set can't be default for the simple reason that DDS apparently isn't supported by OpenSouce Linux drivers, so there are many systems around which don't see any DDS textures at all.

We do in fact have now the regional texture set as default, which is, as the name indicates, even regionally matchable, and for which I spend a lot of time blending textures correctly - so I would appreciate any more specific mismatch complaints, because we can really fix them locally and make a city in Asia look different from a city in the US.
Thorsten
 
Posts: 11086
Joined: Mon Nov 02, 2009 8:33 am

PreviousNext

Return to Hangar talk

Who is online

Users browsing this forum: No registered users and 1 guest