Re: [Flightgear-devel] Aircraft center

[James T. <zak...@ma...>]

> On 10 Jun 2016, at 16:36, Thorsten Renk <tho...@sc...> wrote:
> 
> I understand usually launchers are just fancy ways of assembling a  
> commandline and some of them are just scripted (Python I guess?) - so  
> rather well separable. They all (except FGRun) have maintainers, so there  
> isn't any actual maintenance load for anyone who's not interested.

Well this is the crux point - FGRun complicates the build / install process considerably because we have (up until now) shipped it by default on Windows, but it uses a slightly esoteric toolkit and has no maintainer. My personal preference would be to drop FGRun in our release and nightly builds, and see if anyone wants to pick up maintenance of it as a separate installable, or let it die in favour of one of the other advanced launcher options.

(I’m also happy to discuss an advanced mode for the Qt launcher, if anyone was interested)

But, I have the feeling (reinforced by your poll results) that there’s still a considerable user base for FGRun. What I can't tell is why - if it’s habit (in which case people could move); missing features compared to the Qt launcher (but which ones?) or some other problem. As ever the only way to get large amounts of feedback seems to be intentionally breaking stuff and see who complains loudly, and that is not a great metric anyway.

> It actually has been mentioned here (by yours truly) as a lightweight  
> option replacing PUI for those who don't want to go for heavier Qt option  
> and don't have the hardware to make Phi viable. It's not intended to  
> compete with your Qt efforts in any way, it's for those who don't want  
> additional dependencies.
> 
> I wouldn't consider it as default option (I much agree with your  
> philosophy here), but Qt isn't universally loved by (semi-)power users.  
> Canvas is existing technology which needs to be maintained anyway, thanks  
> to ThorstenB Nasal modules can be loaded 'on demand' and several people  
> are interested in working on it, so I neither see a big maintenance  
> problem nor a performance hit for users which don't need it.
> 
>> So the criteria for a new GUI is ‘better than PUI and with a clear path  
>> to being /much/ better than PUI, both visually and in terms of fixing  
>> long-standing usability bugs’.
> 
> I think that's a good requirement for whatever becomes the default GUI,  
> but for something that's just a lightweight replacement for PUI so that  
> this can go? Novice users don't even need to see it.
> 
> I've heard a lot of unhappy talk in forum conversations and via PM about  
> how the aircraft center was declared 'dead' a while ago - it's not a  
> sentiment shared by everyone. I don't see how keeping a third GUI option  
> open hurts anyone.

The problem is that’s exhausting and limits other transformations of the software, because there’s more things to keep working when migrating to new technologies or systems. (Especially in the rendering / display area) Personally I would rather have /one/ in-sim GUI (plus the option of no-gui and Phi), and I don’t have any desire to spend my time working on a Qt-based one if someone else cares about making a *good* Canvas-based one. It’s just not that interesting a problem for me, compared to working on other areas of the sim. 

I would be delighted if someone is serious about working on the Canvas-based UI, replacing PUI and so on. If there’s people really interested in that, I would much rather help them out, than duplicate their efforts in my limited spare time. Again, I’ve not seen any discussion or code about these approaches proposed here, apologies if I’ve missed it. 

Kind regards,
James