Board index FlightGear Development New features

mp-patch introduction, development

Discussion and requests for new features. Please note that FlightGear developers are volunteers and may or may not be able to consider these requests.

mp-patch introduction, development

Postby jano » Fri Mar 29, 2013 5:58 pm

Hi, from some time, I'm developping more or less in secret :) a patch to compensate lag and reduce the rubber band phenomen: the wiki page

I'd like to have it someday included in FG, and here are my questions about that:

Is it something having enough interest to be included some day?

- I need to have coherent indicated speed, acceleration or angles between the different fdm, this need the bug 202 and bug 901 to be addressed, and i fear those suffer a lack of interest from anyone but me :( .

-currently we are only 2 to test the patch, it would be fun to have more pilots, we can provide windows binary, and patches for linux and mac users, so feel free to ask for a test flight :) (see the videos on the wiki page)

- It would be good to have a ping value to the server, currently we ping before starting fg, and put the value in a nasal file. a minimal ping value from 4 ping every minutes would be good enough for me, but i really don't know how to implement this.

If you know the shotest way to make it appear one day in FG, or want to help in the development, any help is welcome :)

jano

-
jano
 
Posts: 175
Joined: Thu Nov 29, 2007 11:32 pm
Location: france
Callsign: jano
Version: git
OS: debian SID

Re: mp-patch introduction, development

Postby F-JJTH » Fri Mar 29, 2013 6:49 pm

Hi jano,

I'm convinced that your mp-patch is useful. The multiplayer feature seems to be largely popular (looking at the mpmap we have regulary ~60 users) so I think it's important to improve this part of the simulator.
Unfortunately I'm like you : I haven't the magic key for pushing in FG repo :P

It could be good to find a solution for discussion started on the bug tracker (#202 & #901)

Cheers,
Clément
User avatar
F-JJTH
 
Posts: 697
Joined: Fri Sep 09, 2011 11:02 am

Re: mp-patch introduction, development

Postby Hooray » Fri Mar 29, 2013 8:23 pm

I didn't look at your code, and didn't check which projects are affected (fgfs & fgms ?) - but in general, it would seem like a good idea to get in touch with the people who previously committed changes to the affected files. If this affects fgms, you will want to get in touch with the fgms developer obviously - and regarding fgfs, you can check the git history.
Note that these people will probably be pretty busy, but even just getting their signed-off-by may help to have other/active developers show willingness to review/discuss/commit your changes.

That being said, we've discussed a bunch of sensible MP improvements over the years - but HLA is gaining more and more traction, so that people are obviously hoping for the existing MP implementation to be depreciated and replaced with a more flexible HLA-based implementation. On the other hand, this also meant that a number of good ideas didn't end up being implemented, out of fear of wasting our time because of all the HLA work...

Overall, the main issue concerning changes to the existing MP code will be maintaining compatibility with the old system - so if your changes are backward compatible, you should definitely say so - if they are not, it will be REALLY difficult/painful to convince people, especially given the inertia in the MP department.

Generally, the question is how long it will take for HLA to become a viable alternative - I have a bunch of experimental branches here that extend AndersG's binary generic protocol, so that PUBLISH/SUBSCRIBE could be supported in a fashion similar to how the telnet protocol supports both using "onChange"-listeners, a feature I needed for a different project.
Which basically means that the existing MP protocol could be implemented on top of the generic protocol, and the (unmaintained) C++ code could be yanked. On the other hand, in comparison to HLA, all of such attempts would obviously be really ugly hacks :D

To state the obvious, HLA has been frequently brought up during such discussions, but it hasn't yet really materialized to really be a feasible MP replacement at the moment - and I do know, that F-JJTH is also affected by this "HLA movement", because of his fgais code.

Finally, I would suggest to file a gitorious merge request and use the issue tracker to log a review request, that should get you some developer attention (hopefully), and is more future-proof than using the forum or the devel list: https://code.google.com/p/flightgear-bu ... %20request
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: 11186
Joined: Tue Mar 25, 2008 8:40 am

Re: mp-patch introduction, development

Postby jano » Sat Mar 30, 2013 10:55 am

Hi Hooray,
so if your changes are backward compatible, you should definitely say so


they are, i just provide a way to optionnaly display mp planes in the futur, aiming to see others in the same position as seen on the mpserver.
this way our respective position are coherent between players, and an observer would see us is the same respective positions, even if he don't use the patch. (that is only if we are on the same mpserver, having this working on different mpserver would need more work)

it doesn't touch fgms, and is compatible with not patched mp player, except they have a biased velocity sent for yasim plane (bug 202) and both yasim and jsbsim plane don't sent acceleration in the mp paquet (the field exist but is allways 0).

this give some jumpy plane if you display them in the futur, but they are fine otherwise.

About HLA, it really has no importance how i got the mp planes informations, HLA or the current mp protocol, but the plane position prediction will allways be needed to compensate for the lag (except if HLA provide this, i read the HLA doc the same way you read my code :) )


I didn't want to fill a merge request before the bug were addressed, but i can reconsider my position and do a merge request with the bug correction included ...


jano
jano
 
Posts: 175
Joined: Thu Nov 29, 2007 11:32 pm
Location: france
Callsign: jano
Version: git
OS: debian SID

Re: mp-patch introduction, development

Postby Johan G » Sat Jul 05, 2014 10:35 pm

Bump

It seems issue 202 is fixed since Oct 2013. :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)
Johan G
Moderator
 
Posts: 5294
Joined: Fri Aug 06, 2010 5:33 pm
Location: Sweden
Callsign: SE-JG
IRC name: Johan_G
Version: 3.0.0
OS: Windows 7, 32 bit

Re: mp-patch introduction, development

Postby onox » Mon Jul 21, 2014 6:00 pm

Can this patch be applied to one of the servers? Then those who are interested can connect to this server and see the results.
onox
Retired
 
Posts: 431
Joined: Fri Jun 20, 2014 2:45 pm

Re: mp-patch introduction, development

Postby jano » Mon Aug 18, 2014 4:24 pm

Hi there, back after a fg break :)

Johan yep i know, i pushed it :D
the truth is it works quite well here, but not good enough to be pushed in FG yet (from my point of view): a bit difficult to adjust, and using the plane VRP instead of the CG for the rotation center, and some more problems.

for onox, it's not something aplied to a server, but on your fg, I can still provide linux or mac patch if you want to compile it yourself and give a try, or a windows binary if asked, once i manage to plug again my windows HD.

don't expect something too soon, as it's still hollidays :D and as the weather seems good for paragliding :P
jano
 
Posts: 175
Joined: Thu Nov 29, 2007 11:32 pm
Location: france
Callsign: jano
Version: git
OS: debian SID

Re: mp-patch introduction, development

Postby onox » Thu Dec 04, 2014 5:20 pm

Any updates on this? Would be nice to have this in 3.4.
onox
Retired
 
Posts: 431
Joined: Fri Jun 20, 2014 2:45 pm

Re: mp-patch introduction, development

Postby jano » Sun Jun 07, 2015 9:51 pm

Back to the topic after some time on pause mode :D

the good news is that most of the FG pilots use a recent FG with a correct yasim ground speed, the result were bad with earlier versions and some moderate to strong wind.

A basic version of the patch was sent as merge request recently, adding the basic flightgear code, and some nasal to configure the patch's use.
it allow you to display the other planes in the futur, if they are close enough (a distance slider is included).
for now it's up to the user to play with the slider to have the better mp experience (i never managed to gather 3 pilots to test this) and it's better to use the same mpserver.

a good start is to try it with yourself as mp plane (using 127.0.0.1 as mpserver) and play with the menu, if the flightgear part is merged someday ;)

more improvements will follows, but maybe too late for the feature freeze.

jano

ps: if you want to help improve the patch, flying skills are needed to try this on formation flight, refueling :) , the only requirement being that you can compile a patched FG .
jano
 
Posts: 175
Joined: Thu Nov 29, 2007 11:32 pm
Location: france
Callsign: jano
Version: git
OS: debian SID

Re: mp-patch introduction, development

Postby onox » Sun Jun 07, 2015 11:31 pm

I suggest to send a message to the mailing list, because they don't really monitor the merge requests page on SF.
onox
Retired
 
Posts: 431
Joined: Fri Jun 20, 2014 2:45 pm

Re: mp-patch introduction, development

Postby jano » Mon Jun 08, 2015 2:12 pm

the code went in FG, remember that to use it, you need to enable it in the multiplayer=> lag adjust menu:

-the master swich has to be on
-the apply to close aircrafts has to be on too
the slider value for the lag correction is linked to the mpserver ping mostly, and to fps, so i let you try and report.

jano
jano
 
Posts: 175
Joined: Thu Nov 29, 2007 11:32 pm
Location: france
Callsign: jano
Version: git
OS: debian SID

Re: mp-patch introduction, development

Postby Johan G » Mon Jun 08, 2015 2:30 pm

You may want to update the wiki page, Mp-patch, and possibly move it to a more descriptive name, say "Lag jitter compensation in multiplayer" (or something much better :wink: ).

If you move the page, please leave a redirect (otherwise all links to the article will rot). :wink:
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)
Johan G
Moderator
 
Posts: 5294
Joined: Fri Aug 06, 2010 5:33 pm
Location: Sweden
Callsign: SE-JG
IRC name: Johan_G
Version: 3.0.0
OS: Windows 7, 32 bit

Re: mp-patch introduction, development

Postby Necolatis » Mon Jun 08, 2015 3:14 pm

I just downloaded the latest win nightly and pulled latest fgdata.

When playing around with the lag dialog, I get this:

Code: Select all
initialising the mp lag system
Nasal runtime error: undefined symbol: lag
  at d:/ProgramFiles/Development/Git/fgdata/Nasal/lag_adjust.nas, line 31
"Airplane travel is nature's way of making you look like your passport photo."
— Al Gore


Hangar: https://sites.google.com/site/fghangar/
User avatar
Necolatis
 
Posts: 1872
Joined: Mon Oct 29, 2012 12:40 am
Location: EKOD
Callsign: Leto
IRC name: Neco
Version: 2018.2.2
OS: Windows 10 Pro

Re: mp-patch introduction, development

Postby jano » Mon Jun 08, 2015 3:47 pm

arg, i forgot to declare lag, add a line in the top of the file:
Code: Select all
var lag = 0.0;

that should work but not the spectator mode, i surely messed the code when i simplified it
jano
 
Posts: 175
Joined: Thu Nov 29, 2007 11:32 pm
Location: france
Callsign: jano
Version: git
OS: debian SID

Re: mp-patch introduction, development

Postby Tomaskom » Mon Jun 08, 2015 6:27 pm

I just pulled and compiled latest changes to test this, my findings:

-Got the "initialising the mp lag system" message in the console twice upon startup once it remembered it being turned on from previous session.
-Dragging the "Apply to close mp" slider makes no change in the reported number, both in the dialog and /sim/multiplay/lag/range - if I manipulate this property, it changes the number, but slider doesn't seem to read or write it at all. It gets refreshed only when I close and re-open it.
[EDIT]: changing the line 118 in gui/dialogs/lag-adjust.xml from "<object-name>range</object-name>" to "<object-name>Range</object-name>" fixes it - simple case mismatch.
-(Being a perfectionist) please make the first letters of the words in the "lag options" menu entry capitals, like all other menu entries have it ;).

When turned on but without any manual lag adjustments, when I perfectly aligned two aircraft on one FG instance, it got me to about one aircraft length (~13m) of difference on the other instance while flying at 320kt. Very nice result :).

I can assist you with testing, I can bring two or more aircraft at the same time running multiple FG instances, however all but one have to be on autopilot obviously.
"There are no problems in the air. The only problem is hitting the ground"

Get my L-159 ALCA from the FGUK hangar. If you crash it, things gonna go boom!
User avatar
Tomaskom
 
Posts: 132
Joined: Sun Dec 02, 2012 8:03 pm
Location: Czech Republic
Callsign: OK-TomK
Version: git
OS: openSUSE (Linux)

Next

Return to New features

Who is online

Users browsing this forum: No registered users and 3 guests