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 Philosopher » Wed Dec 12, 2012 3:57 pm

Bjoern wrote:These (quite) specific goals can also help involving people willing to contribute as they formulate a very specific task that needs to be tackled instead of throwing about very fancy and very general terms.

I completely agree with this! Just saying "better liveries for every aircraft" isn't going to help, but aye, if someone says pink liveries then we will have an idea of what is required, so that it has a chance of being picked up in someone's spare time (if they agree with it) versus absolutely no chance at all. This needs to be put somewhere, Bjoern ;).
Philosopher
 
Posts: 1593
Joined: Sun Aug 12, 2012 7:29 pm

Re: Help needed - market research for FG

Postby Hooray » Wed Dec 12, 2012 4:06 pm

EigerSA wrote in Wed Dec 12, 2012 3:52 pm:The two posts above have done more to clarify FlightGear for me personally than anything else.


Thanks for letting us know, that's definitely valuable feedback.
In fact, we may even consider making this info more prominently available, or even just get some of the long-term contributors to add the link to their signature, so that newcomers will inevitably stumble across it :D
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: Help needed - market research for FG

Postby i4dnf » Wed Dec 12, 2012 4:18 pm

Bjoern wrote in Wed Dec 12, 2012 3:39 pm:
Surely a tool like NConvert can be fetched during installation and run during startup, right?



No, for various reasons (this is not the place to expose them). This is getting off topic anyway, and the issue has already been discussed, and sofar decided upon some place else.

Don't get me wrong, I'm not against using .dds. It has even more advantages than those presented here sofar. In fact I'm the one that worked on the .dds textures, they were never intended to be more than a "proof of concept" of possible methods of using pregenerated mipmaps to increase/vary visual detail, people assumed otherwise...

FWIW for the planes in my signature I'm using almost exclusively .dds textures due to their advantages, being forced to use a couple .png strategically, due to some bugs in OSG's handling of transparency in .dds
i4dnf
Retired
 
Posts: 743
Joined: Wed Sep 09, 2009 8:17 am
Location: LRBS
Callsign: YR-I4D
Version: GIT
OS: Gentoo Linux ~amd64

Re: Help needed - market research for FG

Postby Hooray » Wed Dec 12, 2012 4:25 pm

The people suggesting the switch to DDS should really up on DDS. The origin of DDS is DirectX, it's a Microsoft technology that uses patented algorithms. FlightGear is open source, cross platform software that uses OpenGL. DDS only really became an option for use in OpenGL due to GLSL (shaders) ARB extensions.

http://www.modwiki.net/wiki/DDS_(file_format)

If you are interested in learning more about the technical and legal reasons for NOT using DDS, please search the mailing list archives or the issue tracker. There are people who are intimately familiar with FlightGear's rendering subsystem and cross platform nature:

https://code.google.com/p/flightgear-bu ... 0Milestone
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: Help needed - market research for FG

Postby i4dnf » Wed Dec 12, 2012 4:33 pm

Hooray, problem is not the .dds format itself, let's stick to facts...
The problem is the compression algorithms (aka compressed .dds, as opposed to uncompressed ones), without which the .dds format looses all of its advantages (bar one, the pregenerated mipmaps that lead to "instant" loading of the texture, but the .dds would inflate to about 1.6x the original uncompressed bitmap size, and yes, that's the reference used, because that's the amount of video ram a texture eats, unless it's compressed .dds which allows for anywhere from 4 to 8 times less memory usage for the same texture, as it stays compressed, and [almost] any GPU is able to read it that way)
i4dnf
Retired
 
Posts: 743
Joined: Wed Sep 09, 2009 8:17 am
Location: LRBS
Callsign: YR-I4D
Version: GIT
OS: Gentoo Linux ~amd64

S3TC (offtopic)

Postby Hooray » Wed Dec 12, 2012 4:49 pm

i4dnf wrote in Wed Dec 12, 2012 4:33 pm:Hooray, problem is not the .dds format itself, let's stick to facts...The problem is the compression algorithms


Well, here's what I wrote:
Hooray wrote in Wed Dec 12, 2012 4:25 pm:The origin of DDS is DirectX, it's a Microsoft technology that uses patented algorithms.


http://en.wikipedia.org/wiki/S3_Texture_Compression
wikipedia wrote:The patent-encumbered status of S3TC and its wide use in software, despite non-encumbered alternatives, have led to a de facto requirement for OpenGL drivers to support it and present a major obstacle to open source implementations.
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: Help needed - market research for FG

Postby i4dnf » Wed Dec 12, 2012 5:01 pm

In your original post you're still "attacking" the format after you state that, instead of focusing on the real issue, which is the compression algorithms...

P.S: You can safely use uncompressed .DDS, without any patent violation. (M$ might have done some good things too, it's not all bad and evil...)
i4dnf
Retired
 
Posts: 743
Joined: Wed Sep 09, 2009 8:17 am
Location: LRBS
Callsign: YR-I4D
Version: GIT
OS: Gentoo Linux ~amd64

Re: Help needed - market research for FG

Postby Thorsten » Wed Dec 12, 2012 5:21 pm

It would be more helpful to keep updated short-to-mid-term to-do-lists for each area, i.e. "Unify shaders", "improve water rendering" in the "graphics" section, "support SID/STAR approaches" in the "ATC/AI" section, "pink liveries for all default aircraft" in "aircraft" etc...
These (quite) specific goals can also help involving people willing to contribute as they formulate a very specific task that needs to be tackled instead of throwing about very fancy and very general terms.


I nowadays usually post 6 months worth of intentions on the devel list after each release to indicate what's next and try to evaluate it during the discussion prior to a release about 5 months later. Major projects (Rembrandt, Atmospheric Light Scattering, Canvas,...) have forum threads to illustrate the progress. The newsletter usually contains info about what is currently going on and what is planned.

It's not that the info would not be available, there's just not the one place where everyone summarizes. If we had such a central a place, I wouldn't mind posting it there, but I really don't buy into the argument that a moderately interested forum user could not know what I'm up to. If anyone is *really* interested in contributing to weather, it probably takes less than 5 minutes to find out who should get the PM. Likewise, I don't think it's outrageously difficult to get in touch with, say, terragear people.

I mean, it's probably nice to have a summary paper of anything, but the relevant people are actually around, and simply asking them is a valid strategy as well if you really want to get involved. I really don't mind helping someone else getting a project on the way, but for that I can specifically assign time, whereas otherwise I am asked to spend time to create the infrastructure 'just in case' - which has a lot of loss if the case doesn't arise (suffice to say, my experience in that department wasn't exactly hopeful - about 80% of the projects where I started coding something on request simply disappeared).
Thorsten
 
Posts: 12490
Joined: Mon Nov 02, 2009 9:33 am

Re: Help needed - market research for FG

Postby Gijs » Wed Dec 12, 2012 7:05 pm

Just a small request to everyone, please think about the time it takes to write all these lengthy posts, and compare that to the time it would take to actually fix most of the things you are discussing... :shock:

I think some of you are far better in discussing that, so just a few short remarks of mine on sentences that caught my special attention.
EigerSA wrote:Reading post after post of the apt.data thread and trying to figure out how it all works only to accidentally stumble on a thread mentioning WED is frustrating!

I think we improved this situation a lot, earlier this year, when we revamped the wiki portals. On the main page you go to the "Development > Scenery" portal, where you'll find the link to http://wiki.flightgear.org/Howto:Make_an_airport within the glimpse of an eye. That article summarises everything involved in making/updating an airport pretty well IMHO...

To both Thorsten and Hooray, my apologies, I only came across this now... if I had seen it earlier I probably would have reacted differently.

http://wiki.flightgear.org/How_to_the_F ... ject_works

Please, please remember to do something about it, if it took you a long time to find this. The wiki is open for everyone (after registration) and nothing stops you from adding a link in a clear place. This specific article is now mentioned on the developer portal (thanks Hooray) and the main page as well.

I do read every single post on this forum, but I don't have the time to fix everyone's wiki problems/requests ;-)
Airports: EHAM, EHLE, KSFO
Aircraft: 747-400
User avatar
Gijs
Moderator
 
Posts: 9544
Joined: Tue Jul 03, 2007 3:55 pm
Location: Delft, the Netherlands
Callsign: PH-GYS
Version: Git
OS: Windows 10

Re: Help needed - market research for FG

Postby Michat » Thu Dec 13, 2012 7:44 pm

BJORN For extendability, simplicity and keeping in FG's "everything is done procedurally" spirit, I'd suggest a text-to-speech system (yes, you may cringe), if there is a FOS one. But that's a future issue.


I asked for text-to-speech 3-4 years ago on IRC, our dear hackers guide me to a engine created by some UK University that I don't remember it's name. I good friend of mine 727 retired pilot lost his voice due to cancer, suddenly I understood how important is the accesibility. Without doubt it will be a gem for FG. Our 727 is not a gem, but he was very impressed about how intrepid can be the freesoftware people: Trying to help I give him all the info links, we find nice info, however we are not programers, anyway I don't loose the faith.



In other side, but related with FG promotion, as I commented in my first post:

#4SpreedOurMessage
_______________________________________________________________________________________________________________________________________
Loud, we are part of a culture we add free culture and common sense to our simulator and to your world.
We are not alone FSF - GPL - Blender - GIMP - OPENAL - GIT - mediagoblin - Inkscape
_______________________________________________________________________________________________________________________________________

FSF released "giving guide 2012" https://www.fsf.org/givingguide.
I think is a good opportunity to show us versus FS-X-PLANE privative software model.
We have a lot of advantages versus privatives.

15 years, mature, multiplatform, sofisticated flight simulation software under GPL ETC
The best community. ETC
We are one of the great GPL big games. But we are a simulator not a game. We have presence atn universities, museums, real life events such a linuztag and fsweekend. ETC
FG is a good example ot it. We use Gimp, Inscaspe, Blender. We can use mediagoblin to share 3-D models and we can print those aircraft models using the LuzlBolt 3D printer. We can connect to arduino, android, and to Mathlab (wishing for a free solution). We are friends of emacs editor. ;)

We have a charity program. And of course a better assistance than privative competitors.


ETC ETC.
Please can anyone help , requesting and improving a short versus (comparative) article for this great initiative?
Will help a lot doing fgembassy program reliable. Because we are not alone


http://libreplanet.org/wiki/Group:Giving_Guide_2012
I'm running a 386, I can't do more.
Thanks.
Last edited by Michat on Thu Dec 13, 2012 9:59 pm, edited 1 time in total.
User avatar
Michat
 
Posts: 1226
Joined: Mon Jan 25, 2010 7:24 pm
Location: Spain
Version: 191b
OS: MX 21 Fluxbox oniMac

Re: Help needed - market research for FG

Postby Hooray » Thu Dec 13, 2012 8:17 pm

You are probably referring to the "Festival" TTS engine, which is still supported by FG.
There are solutions for turning speech into text, such as Sphinx.
BTW: I feel the other GPL/freedom related issues you are raising are valid, but should be better moved to a separate thread (possibly even to the wiki/newsletter), no?
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: Help needed - market research for FG

Postby Michat » Thu Dec 13, 2012 10:18 pm

Hooray wrote in Thu Dec 13, 2012 8:17 pm:You are probably referring to the "Festival" TTS engine, which is still supported by FG.
There are solutions for turning speech into text, such as Sphinx.
BTW: I feel the other GPL/freedom related issues you are raising are valid, but should be better moved to a separate thread (possibly even to the wiki/newsletter), no?


Yea Festival good memory , I remember now also CIgWin, Don't know about Sphinx. But how to PTT over FGCom? HUge task.

Agree with you and happy you wellcome this issue GPL/Freedom. Agree with the wiki newsletter also. However time is running, we must first write a good versus short article and request it on libre planet. Then we can translate it to other languages in order to push hard.

I'm so so sorry, here with my 386 I'm having a lot of difficults even to connect trough www. Now I'm trying to edit with out to see what i'm writting, so imagine.

Please help
User avatar
Michat
 
Posts: 1226
Joined: Mon Jan 25, 2010 7:24 pm
Location: Spain
Version: 191b
OS: MX 21 Fluxbox oniMac

Re: Help needed - market research for FG

Postby Johan G » Fri Dec 14, 2012 10:15 pm

Michat wrote in Thu Dec 13, 2012 7:44 pm:We can connect to arduino, android, and to Mathlab (wishing for a free solution).

GNU Octave is a relatively compatible open source competitor to Matlab (MATrix laboratory, not MATHs laboratory), though with probably a lot fewer ready to use simulation packages.
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: Help needed - market research for FG

Postby Bjoern » Sat Dec 15, 2012 2:30 am

DDS:

I'll just supply a "download nconvert and convert .dds files script" in the next iteration of my FG-from-GIT-bash script...no licensing issues and an improved user experience for those using open source drivers.

DDS files are present in FG and are being widely used (judging from the "switch to materials-dds" posts in the forum) after all.



Thorsten wrote in Wed Dec 12, 2012 5:21 pm:I mean, it's probably nice to have a summary paper of anything, but the relevant people are actually around, and simply asking them is a valid strategy as well if you really want to get involved.


Well, a centralized information document (wiki page; also containing a link to and note about the issue tracker) that states the forum handle of the person involved would fulfill both, wouldn't it?



Philosopher wrote in Wed Dec 12, 2012 3:57 pm:This needs to be put somewhere, Bjoern ;).


1. Print out
2. Frame
3. Put up on wall
4. ???
5. PROFIT!

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

Re: Help needed - market research for FG

Postby i4dnf » Sat Dec 15, 2012 6:16 am

Bjoern wrote in Sat Dec 15, 2012 2:30 am:DDS:

I'll just supply a "download nconvert and convert .dds files script" in the next iteration of my FG-from-GIT-bash script...no licensing issues and an improved user experience for those using open source drivers.

DDS files are present in FG and are being widely used (judging from the "switch to materials-dds" posts in the forum) after all.


Please don't. A straight, blind, conversion is not possible. Some of those .dds files contain manual mipmaps ( The mipmaps are done by hand to present you a different texture at each miplevel). The others are just the current .png textures flipped vertically then converted to .dds. There are other issues too.
Converting the .dds files to .png will not improve anyone's experience, it might actually worsen it...
Converting from compressed .dds to uncompressed .dds is not an option either, it will blow up (at least 4x up to 16x) memory usage (especially Video RAM usage)
Also it won't bring those users anything extra, that they don't already have available, given the limitations of the oss drivers.

There is a more viable solution by using libtxc_dxtn (provides mesa with the texture compression extensions), so you could hint at installing that, and probably an extra step of enabling texture compression in xorg.conf is needed (this varies with drivers, I'm not familiar with the exact steps).
i4dnf
Retired
 
Posts: 743
Joined: Wed Sep 09, 2009 8:17 am
Location: LRBS
Callsign: YR-I4D
Version: GIT
OS: Gentoo Linux ~amd64

PreviousNext

Return to Hangar talk

Who is online

Users browsing this forum: No registered users and 1 guest