[Flightgear-devel] Technical tasks for 2018.3

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

Hi,

Since there seemed to be near-universal agreement on making 2018.3 a big release (whatever term we use), I wanted to outline the technical items I’m aware of (and collect any more / remove any unrealistic ones). We also need to discuss the social aspects of the release - and who does them of course.

I’m on vacation from 5th July to 14th July, and won’t be reading emails - I’m thinking we can start the social / organisational discussion after that, but if I come back from holiday and it’s all figured out, even better.

(The following list is what I **think** we could / should get done for 2018.3, it’s up for discussion, and based on what I’ve read here)

Things I know about:

	- the launcher is now feature complete with the QtQuick UI, I’ll keep making bug-fixes of course as people report them, but my todo list for the core features is now done

	- I **might** have time to add one new feature I personally want: launcher flightplan loading / parsing, i.e taking something like http://rfinder.asalink.net/free/autoroute_rtx.php which produces ASCII along the lines of: 
	"EGPH SID TLA L602 OTR L90 DOLAS L603 LAMSO UL603 EVELI UL19 TULIP UM624 DIK UN852 GTQ UZ343 BEGAR UL15 RIPUS UN850 ODINA DCT SRN DCT KALIK DCT GIPIX DCT XIBIL STAR LIRF”

.. and setting up the flightplan inside FG. I’d be curious to hear if anyone besides me would use such a feature, or if there’s other flight-plan formats we could support. (Again, no promises!) (The actual parsing would be done in FG, but the launcher would pick your starting location, etc from the loaded flight-plan)

	- I need to finish off Florent’s translation work so we can do updated translations across the whole UI. 

	- get the translations updated (social aspect to this of course)

	- Stuart and Torsten can hopefully get the scenery-models LOD-stripping and STG-merging done, which should give us a nice rendering performance boost (and faster downloads, and a happier OpenSceneGraph due to fewer nodes...)

	- Tim Moore’s canvas-rendering-on-a-pre-camera fix, again should give us a nice perf win for Canvas-intensive acft

	- Torsten is working on tarball-ing Terrasync directories which should really help terraSync download performance. I will need to do a bit of work to support the new logic inside FG, I know Keith Patterson is planning the same for terra-master

	- Stuart’s MP-model-mapping-table, maybe?

	- improve meta-data (description, ratings, splash screens, tags) on even more aircraft in FGaddon (a community effort, yay!, once we organise it….)

	- some carrier dialog changes from Richard? (But I may be confused about that?)

	- preview mode of Fernando Garciá Linán’s compositor work? (Depending on how he feels about its stability & the time frame!)
	(Or should we leave it out of 2018.3 so it can be worked on during 2019.x?)

Major bugs I know about:

	- crash in comm radios related to iax_client

	- mouse button stickyness in PUI sometimes

Things I don’t know about:

	- updates / fixes to the C172

	- get the FG1000 into an acft, possibly the C182, so we can highlight it as a new feature

	- updates / fixes / bugs to other ‘major’ aircraft we might want to highlight during the release process (I have no idea how we define ‘major’ here, but e.g the Shuttle or 777)

	- default airport (PHNL) improvements / fixes, are there any pending or needed?
	
	- ‘tutorial’ airport changes / additions to FGData

	- Yasim updates (Henning?) and risk profile for those

If we got all of the above done, I think we’d be in nice shape for a release, and some of the above are ‘nice to have’, not critical path.

So, what have I missed? For bugs, I would strongly prefer they are in the SF issue tracker, don’t just start spamming this thread. If I’ve presumptuously named you in the list above, it would be good to hear how realistic I’ve been, or if you’d need help to get things done.

If you want to propose technical additions to the list, please consider how to resource them as well - eg for myself what I’ve outlined feels practical, without being overly ambitious. (I need time to work on the Simpit as well, for FSWeekend, so expect to see remote-canvas / HID / performance work also). 

Kind regards,
James