Board index FlightGear Development Aircraft

New 777 3d Cockpit

Questions and discussion about creating aircraft. Flight dynamics, 3d models, cockpits, systems, animation, textures.

New 777 3d Cockpit

Postby I-NEMO » Tue Nov 05, 2013 1:33 am

Hallo FG pilots,

greetings from Italy!...the 777's NEW 3d Cockpit is coming...

Image

I was absent from FG for quite a long time, but I started again my work on the 777 cockpit, as previously reported more than one year ago...
My aim is to try to get a more realistick 777's cockpit; I'm actually working on a pre-release version, and my big thanx goes to Hyde for support and consideration.
I was moved to publish some preview shots while recently reading a post about poor illumination at night (no Rembrandt).
Actually, the new model (an Alpha release, though) won't be ready before one more month (I presume), because of some fine tuning required by proper modeling (Blender) and texturing (Photoshop). The result, in my opinion, is coming out nicely...and I do hope that 777's lovers will like to fly sitting in a more 'warm' cockpit...
Here are some shots from the first pre-release:

ImageImageImageImageImageImageImageImageImageImageImageImageImageImageImageImageImageImage

The model comes with full lightmap and I did not dare to change any of the actual controls of the 777 (names, objects and so far are compliant with the standard official version).
The objective is to get a good graphically rendered cockpit, then I'll move on - under Hyde supervision - to add some more functionality for all the modeled switches and controls...the cockpit's model is almost complete, and I would like to know if someone might help with the weird problem posed by some main-panel instruments' displays, which overlap weirdly with some other objects once rendered in FG...I remember that somewhere on the forum the issue was addressed (resolved?), but cannot locate the correct thread...the issue probably depends from the vieving position, and in my 3d modeler experience is commonly addressed as z-buffer problem, that is, setting the correct priority in 'depth'...

Be as it may, does anyone know what should be done? Hyde tried to address the problem through the new Canvas system, but results are not safe/stable yet...so, before releasing the full 3d Cockpit (for those who will like it) I would like to eliminate that disturbing visual effect: the instruments looks perfect in Blender, and not a single bit has been changed in the .xml or the .ac relevant files (actually, the 3d models are the same used in the existing official version).

Please be advised that the new 3d cockpit is very heavy in terms of Ram and resources, because of the details, the numeber of objects added, and so on.

My configuration:

Intel i7 3770 8 cores
32 Gb Ram ddr3
NVidia Gtx 650 TI
Windows 7
FG 2.12

On this, with the new 3d cockpit fully loaded, I fly nicely over France at 59 fps...and - thanx to the excellent work by Hyde - the AP of the 777 works like a breeze...
I'll be happy to share my humble work, stemmed out from simple passion for the beautiful 777...to the FG community, of course!

Regards

I-NEMO
I-NEMO
 
Posts: 102
Joined: Thu Sep 29, 2011 3:09 am
Location: Italy
Callsign: I-NEMO
Version: 2017.2.1
OS: Windows 7 64 bit

Re: New 777 3d Cockpit

Postby Hooray » Tue Nov 05, 2013 1:44 am

That looks absolutely fantastic! Great to see that you coordinated your work with Hyde!

Hyde tried to address the problem through the new Canvas system, but results are not safe/stable yet...

I saw the corresponding commit message, but I don't believe this to be performance problem of the canvas system, it's probably due to certain Nasal coding patterns - maybe someone could briefly explain what "unsafe/stable" means in the context of a Nasal/Canvas-based avionics stack ?
For example, Gijs' recent 744 work doesn't seem to suffer from performance problems - even though there's quite a bit of optimization potential left coding-wise.
I previously got in touch with TheTom and suggested to write profiling stats for each canvas texture to the property tree, so that we see where time is spent, i.e. for each element/group - but I still believe that even complex MFDs (like the ND of the 744) are likely to primarily have a Nasal overhead, not so much due to the canvas system though.

I haven't looked at the Nasal/Canvas code in question, but I assume that it's simply not yet very optimized - so I'd suggest to make it available again, so that we can have a look and see if we spot anything that's obviously problematic.

But please tell us first of all what exactly the problems were with the reverted code :D

PS: PLEASE do add your news and your screen shots to the upcoming Nov. newsletter (see my signature)
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: 12707
Joined: Tue Mar 25, 2008 9:40 am
Pronouns: THOU

Re: New 777 3d Cockpit

Postby I-NEMO » Tue Nov 05, 2013 2:12 am

Hooray,

thanks for your quick reaction: I appreciate your compliments!

Regarding tha Nasal issue, I'm not too competent on that, and would gladly leave the eventual answer/solution to the experts on the matter.
For the moment, though, I remember that the same visual problem had been reported several months ago in the Forum (possibly even back to two years ago). Since I expect that the actual solution through Nasal/Canvas would take some time - as it is understandable - I'm wondering if that problem had been somehow resolved or not (while Canvas was not around, yet). So the actual new 3d cockpit would not suffer of that very disturbing effect...I may be wrong, but possibly Gijs gave an answer on that: I tried a search on the forum, but - probably because of my age - I was not able to get the right keyword or phrase to get that old thread...

Regarding the November Newsletter: I'll gladly comply with your kind request, but only once I'll have finished a clean-up of the various directories where I added Textures, test, code, and so on (five months of daily work!); by next week, I presume, I'll be able to work on the newsletter...

Thanx again.

Regards,

I-NEMO
I-NEMO
 
Posts: 102
Joined: Thu Sep 29, 2011 3:09 am
Location: Italy
Callsign: I-NEMO
Version: 2017.2.1
OS: Windows 7 64 bit

Re: New 777 3d Cockpit

Postby Hyde » Tue Nov 05, 2013 4:48 am

Hoory,

Sorry I threw ambiguous comment but recently many big change happens simultaneously such as canvas implementation, terrasync directory change, world scenically 2.0 release.
I'm not sure what is related but FG becomes unstable now. I often could not start the program. It crashes without any error messages.
Since canvas implementation is not practical yet I decided to revert back for investigation.
It's much better than before but still sometimes get crash so I think this is not related to canvas.

I-NEMO,

I found the solution of see through issue.
Until canvas becomes perfect, I will tweak current implementation.
Hyde
 
Posts: 724
Joined: Fri Aug 12, 2011 2:09 pm
Location: Peachtree city, GA
Callsign: N407DS
Version: GIT
OS: Fedora 25

Re: New 777 3d Cockpit

Postby jormapaappa1235 » Tue Nov 05, 2013 10:36 am

The cockpit looks simply amazing! I must do better with my upcoming cockpit releases ;) Where have you got that nice textures from?
jormapaappa1235
 
Posts: 160
Joined: Fri Apr 05, 2013 5:52 pm
OS: Windows 7, Ubuntu

Re: New 777 3d Cockpit

Postby Hooray » Tue Nov 05, 2013 1:14 pm

Hyde wrote in Tue Nov 05, 2013 4:48 am:I'm not sure what is related but FG becomes unstable now. I often could not start the program. It crashes without any error messages.


Hyde, your profile says that you are on Fedora/Linux, and probably building from source ?
If so, you are in an excellent position to provide additional information using a gdb backtrace for example.
I'm sure we can walk you through the whole process, it should not take much longer than 5-10 minutes actually.

In the meantime, you could try to reproduce the issue, note down all settings you are using, and then start FlightGear with --log-level=debug (or even bulk).
Afterwards, go to your $FG_HOME and upload the fgfs.log file to a pastebin so that we can have a look and see what FG was doing before it crashed.

To see if it's scenery related, just start up in a location with little scenery, or without any (sea/ocean).
To see if it's aircraft related, just switch to a less complex aircraft (the ufo should always work).

If the issue persists despite using the ufo in a non-scenery area, it must be some other problem.

By the way, to provide a backtrace - you just need to install "gdb" via sudo yum install gdb and then run gdb, while passing the path to the fgfs executable, for example:

gdb flightgear/src/Main/fgfs

That will load the fgfs binary in the gdb debugger.

Next you can run FlightGear via run (pass any arguments here after run).


And once it is crashing somewhere, just execute this:thread apply all bt full

Note that it is beneficial to have a FlightGear debug build in this case, because there will be more information available that way.
A debug build can be easily created during cmake configuration: http://wiki.flightgear.org/Building_usi ... bug_Builds

Ideally, SG/FG would be configured like this (a osg debug build is rarely needed).

Usually, you can simply copy your current build directory for sg and fg - for example to fg.debug and sg.debug, and then re-run cmake inside to create debug versions of sg&fg.

This may seem complicated at first, but it shouldn't take much longer than 20-25 minutes - and afterwards, you can easily test things in a debug build.

Finally, do note that debug builds are usually slightly slower than an optimized/release build.
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: 12707
Joined: Tue Mar 25, 2008 9:40 am
Pronouns: THOU

Re: New 777 3d Cockpit

Postby Hyde » Tue Nov 05, 2013 5:15 pm

Hooray,

Thanks for your elaborate advice.

I completely remove the .fgfs directory and restart fg then issue's gone. :D
I will use those next time.

One trivial information for fgrun 1.7.0.
Since terrasync directory is moved under .fgfs, we can not register the new airport in Select a location page.
Once you start with new .fgfs directory, reposition to desired aiport.
Then close flightgear and return back to the first page of fgrun.
Remove current FG_SCENERY: directory then enter /home/your id/.fgfs/TerraSunc
Then Refresh on Select a location page. Now you can select your favorite airport.
Refresh whenever you fly to new place so that keep list is updated.

This might be changed near future.
Hyde
 
Posts: 724
Joined: Fri Aug 12, 2011 2:09 pm
Location: Peachtree city, GA
Callsign: N407DS
Version: GIT
OS: Fedora 25

Re: New 777 3d Cockpit

Postby Hooray » Tue Nov 05, 2013 5:23 pm

Hyde wrote in Tue Nov 05, 2013 5:15 pm:I completely remove the .fgfs directory and restart fg then issue's gone. :D


That's also interesting, and could possibly be related to the built-in navcache - next time, you could try to rename the cache files and see if that has the same effect or not.
BTW: Instead of removing (deleting) things completely, it is better to rename files/folders (i.e. to fgfs.old or fgfs.broken), so that you can more easily revert back to the old data - that would at least help us with troubleshooting.
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: 12707
Joined: Tue Mar 25, 2008 9:40 am
Pronouns: THOU

Re: New 777 3d Cockpit

Postby Gijs » Thu Nov 07, 2013 6:48 pm

jormapaappa1235 wrote in Tue Nov 05, 2013 10:36 am:Where have you got that nice textures from?

Right, interesting question. They look very, very similar to the Project Opensky textures. Did you get permission to use them? And if so, also under the GNU GPL? It would be a shame if we couldn't make use of all your hard editing wor. I can see you've really spend some effort improving those textures and the result looks great...
Airports: EHAM, EHLE, KSFO
Aircraft: 747-400
User avatar
Gijs
Moderator
 
Posts: 9542
Joined: Tue Jul 03, 2007 3:55 pm
Location: Delft, the Netherlands
Callsign: PH-GYS
Version: Git
OS: Windows 10

Re: New 777 3d Cockpit

Postby I-NEMO » Fri Nov 08, 2013 3:26 am

Hallo Gijs,

thanks for the gentle approach of your 'interesting' questions, to which you certainly know the answers already!

The original textures are of course coming from Project OpensSky (POSKY), as they are easily available on the web (together with other aircrafts' ones).
Those textures have been extensively modified in many ways, so to approach the real 777's cockpit (the list of the mods would be quite long to the purpose of this reply, but I'll be glad to enter into details, should you feel it necessary).
POSKY, as it's widely known, closed down some years ago; one of the two groups (ex-members, as it is reported) planned to mantain the POSKY 'heritage' through a new website, which actually never saw the light.
Since more than two years, when I started to work on the Overhead panel in FG, I've tried several times to get in touch with someone who could answer my due request for permission: I've never received an answer. My last EMail to the current maintainer of the POSKY archive (Mr. Aiello) did not receive an answer as well...
I do not know who to contact for the permission's request (whether GPL or not): I should assume (but I may be wrong) that the 'Master Painter' - Mr. Yosuke Ube - is the author of the original work; but I don't know how to get in touch with him...
I do not think that Paul Craig is the owner of the textures I used; in fact, Craig's textures are slightly different in origin from those I've thought to be appropiate for my modeling project.
The EULA of the package I've downloaded is quite ambiguous and difficult in assessing what is allowed and what not...further, as I said, it's certainly not common to contact a website (or else) which currently does not exist anymore.

Be as it may, this is my current standing on this issue:

    The original copyright of the original textures is held by POSKY.
    I do ignore who (i.e., the name of the author of the pictures, and/or of the first 'clean up' of the photographic work which was the basis for POSKY's graphic rendition) is the true original author of the textures I've chosen to use for modification (Since POSKY is not existing anymore, I think that the physical person responsible for the original authorship might be credited for a proper copyright statement).
    I've not received any answer whatsoever to my Emails on the permission/license issue.
    I'm the author of all the modification made to POSKY pictures for my current development project.
    All POSKY's works (not only textures, but models, FDEs, and so on) were not GNU, nor GNU GPL, licensed. They are just freeware, which has nothing to do with the point we're dealing with.
    Once my project will be finished - even at pre-release level - I will of course specify in a file included in the package that the original copyright of the textures by POSKY is POSKY, no matter if POSKY might be 'reachable' or not.
    My own works (mods and originals) will be copyrighted under my name, and I will of course offer my own works as GNU GPL.

This is, I presume, a fair starting position for my project: should you or any other member of the community feel that this might endanger or breach the FG's policy, I won't complain if the new 777's cockpit will not be included in any release of FG; I will hold the model for my personal pleasure, flying on it with my copy of FG. Because that's the reason I try to fly with a flight simulator: just pleasure.
Should anyone of FG's community ask for the model, if he/she likes it, I'll set up a downloadable link somewhere in the future, specifying that the original textures I used for my project were/are by the POSKY...

Hope this might be considered as my best effort to comply with copyright issues: we all have to thanks POSKY's folks for the great job they have made for the flight simulation world (freeware and payware), and at the same time I'm well aware of the long, long, long time I've spent on modeling, painting, fine-tuning the various pieces of the 777's 3d cockpit (Blender as well, not only Photoshop); like a giant puzzzle, it's made for pure passion and hoping that someone else might like to fly in a plane that 'simulates' a plane also visually.
If I ever had a chance to spend three full days (and nights) in a real 777's cockpit with a good Nikon, I would have certainly chosen my own pictures to draw, paint, resize, filter and tune up much better textures than those done some years ago by the POSKY's team (which are in many cases unprecise, if not unfitted and/or badly projected).

If you feel that
It would be a shame if we couldn't make use of all your hard editing wor.

well...I did my best effort, because most of the 3d modeling too has been made by scratch, also on 90% of pre-existing 3d objects of the old FG's 777.
[Note: while 3d modeling with Blender, I discovered that most of the cockpit had wrong measures and proportions, and misplacements, even by bare comparison of true 777's pictures with the actual FG 777's cockpit]

I do of course like the GNU GPL choice made by FG's team: I do think, however, that most of the actual aircrafts' models are missing the point of visual pleasure because of a number of reasons. And that's a real pity.
Please, rest assured that it's just my opinion, of course.
Should you feel to share an advice on this issue, I'll be glad to read your notes.

Regards,

I-NEMO
I-NEMO
 
Posts: 102
Joined: Thu Sep 29, 2011 3:09 am
Location: Italy
Callsign: I-NEMO
Version: 2017.2.1
OS: Windows 7 64 bit

Re: New 777 3d Cockpit

Postby Johan G » Fri Nov 08, 2013 12:35 pm

@ I-NEMO:

That Is the most impressive and well thought through answer I have to date seen to a question about the license and source to something appearing on the forum. It is a pleasure to see that you have tried to get in contact with original authors as well as are being very aware of both copyright and license issues. :D

Even if you can only share screenshots (at least for now), they are still inspirational. And just maybe someone can get around to help you with the photos you need of the cockpit in order to make completely new, and more accurate models and textures. :wink: :D
Low-level flying — It's all fun and games till someone looses an engine. (Paraphrased from a YouTube video)
Improving the Dassault Mirage F1 (Wiki, Forum, GitLab. Work in slow progress)
Some YouTube videos
Johan G
Moderator
 
Posts: 6625
Joined: Fri Aug 06, 2010 6:33 pm
Location: Sweden
Callsign: SE-JG
IRC name: Johan_G
Version: 2020.3.4
OS: Windows 10, 64 bit

Re: New 777 3d Cockpit

Postby Bjoern » Sat Nov 09, 2013 7:38 pm

As long as you give credit to the original authors (e.g. Mrs Yosuke Ube) and are willing to deal with any possible inquiries from the original authors (who are, I think, in a grup named TDS now), I don't see much of a licensing problem.
Yet, I would not license the model under GPL as it includes other people's work that was released undera different license.
In terms of FG, this means that the model can't be put into the official FGData package, but I don't see a problem with making it available for download elsewhere*.


Anyway, thanks for putting so much thought into licensing.


*I really think there should be a "non-GPL content" repository for FG.
Bjoern
 
Posts: 484
Joined: Fri Jan 06, 2012 11:00 pm
Location: TXL (RIP)
Version: Next
OS: ArchLinux

Re: New 777 3d Cockpit

Postby I-NEMO » Sun Nov 10, 2013 2:42 am

Thanks for your replies; of course I'll credit POSKY at general for the original work, and include all the names mentioned in the accompanying EULA.
I leave to the experts the decision about an eventual "non-GPL content" repository. Indeed, it looks improbable that most of the photography-based work available (now and in the future, for aircraft, or terrain, or else, connected with true
realistic-graphics issues) could fall within the strict GNU GPL's boundaries.
Again, that's just my humble opinion.

By the way, Bjoern, I did not know that Yosuke Ube is a 'Mrs' and not a 'Mr'...! Thanks for avoiding me a bad fall on 'style'...
My work is going on, with the assistance of the generous and passionate Hyde, and the new 777's Cockpit Model should be available as a first-release in about a month (I hope).

Regards,

I-NEMO
I-NEMO
 
Posts: 102
Joined: Thu Sep 29, 2011 3:09 am
Location: Italy
Callsign: I-NEMO
Version: 2017.2.1
OS: Windows 7 64 bit

Re: New 777 3d Cockpit

Postby Bjoern » Sun Nov 10, 2013 7:09 pm

I-NEMO wrote in Sun Nov 10, 2013 2:42 am:By the way, Bjoern, I did not know that Yosuke Ube is a 'Mrs' and not a 'Mr'...! Thanks for avoiding me a bad fall on 'style'...


Welcome. ;)
Bjoern
 
Posts: 484
Joined: Fri Jan 06, 2012 11:00 pm
Location: TXL (RIP)
Version: Next
OS: ArchLinux

Re: New 777 3d Cockpit

Postby Gijs » Sun Nov 10, 2013 7:27 pm

Sorry for my delayed reply, was away most of the weekend. I'd also like to apologise if my post seemed sarcastic or dismissive. That was certainly not my intention. I did not know the answers to the questions I've asked, else I wouldn't have asked them.

As said, my intention is not to criticise your work or have it removed from anywhere. But we have to respect licenses. We like people to respect our licenses (we don't want people to mis-use our hard work); we should also respect other people's licenses. The fact that the 777 is in our base package shouldn't really matter (and does not legally), but it means this aircraft serves as an example for other developers and as first impression for many. We therefore need to be very clear that it complies with the GPL.

Copyright laws don't care whether you've done your best contacting the original author(s) or not. You are simply not allowed to re-use something in a way other than permitted by the license it was released under. There's no such thing as "silence gives consent". In your case that's very unfortunate (as the authors are gone), but at the same time it's a welcome protection that also protects your work...

I did my best effort, because most of the 3d modelling too has been made by scratch, also on 90% of pre-existing 3d objects of the old FG's 777.

Then we can still use those models, right? It's "just" the textures that aren't GPL compatible. Please do try to coordinate that with Hyde and see if we can have at least those nice models included in Git.

Cheers,
Gijs
Airports: EHAM, EHLE, KSFO
Aircraft: 747-400
User avatar
Gijs
Moderator
 
Posts: 9542
Joined: Tue Jul 03, 2007 3:55 pm
Location: Delft, the Netherlands
Callsign: PH-GYS
Version: Git
OS: Windows 10

Next

Return to Aircraft

Who is online

Users browsing this forum: No registered users and 2 guests