Re: [Flightgear-devel] recent changes to navdisplay?

[James T. <ja...@fl...>]

> On 8 Apr 2020, at 15:34, Tobias Dammers <tda...@gm...> wrote:
> 
> On Wed, Apr 08, 2020 at 02:27:45PM +0100, James Turner wrote:
>> 
>> The other thing I hadn’t realised until reading the description again,
>> is that this is not the Canvas based ND, but something else - so what
>> / how does it work? My first through was to check the ND in my 737 but
>> of course that is working fine. Is it using the old C++ ND code I
>> wrote pre-Canvas? Or something else again?
> 
> Yes, it's the pre-Canvas C++ NavDisplay implementation. Not sure who
> wrote it, but I think there are a few other aircraft models (Citation?)
> that use it. It works by implementing a 2D panel with XML, and then
> referencing that in the 3D model, mapping it onto a 3D surface.

Hmm, this is a little strange. I wrote the pre-Canvas C++ NavDisplay, but that definitely does /not/ use 2D panels. It works like the WXR-radar and some other ‘OD gauges’ from a long time ago : they appear in the 3D scene as a dynamic texture. (Referenced by looking up a dummy texture name, usually present as a white placeholder in the files on disk). 

As far as I know, there is zero involvement in the 2D panel code, but of course it’s probably common to combine the NavDisplay with some 2D panel elements (eg, text). 

Anyway, I’ll do some archeology and try to figure out What The hell Is Going On (TM)

I would also strongly recommend switching to the Canvas NavDisplay at this point, the C++ one is never going to receive any maintenance or updates. I see you mentioned an issue with Canvas on multiple-monitors, but I don’t think that is a systematic issue - you should probably open a ticket about your setup and start debugging what’s going on there, since at this point Canvas is basically a required component.

Kind regards,
James