STAGE_PRIORITY Delta-V calculation

[Archgeek]

So having done a bunch of work on a couple of puller-type ion craft designs featuring too many drop tanks, I've been reminded that neither major dv-calculating mod quite has a grip on dealing with drop tanks in STAGE_PRIORITY flow (xenon and mono) and will woefully underestimate a craft's delta-v, as seen here:

Older screenshots, but I've confirmed it's not been fixed yet.

I dug around the KER thread last week but couldn't find anything new regarding it since during the open beta, when I helped @Padishar isolate the issue -- so I must inquire: where precisely should I look or ask regarding the status of STAGE_PRIORITY fuel flow delta-v calculation?  I sense it's probably an annoyingly complex problem with too many edge cases, and I don't want to just derp into the mod's thread asking for a progress update on a specific feature like a jerk.  Are there defined channels for this sort of thing, or tracker I can follow?

[Padishar]

8 hours ago, Archgeek said:

I don't want to just derp into the mod's thread asking for a progress update on a specific feature like a jerk.  Are there defined channels for this sort of thing, or tracker I can follow?

Well, you wouldn't be considered a jerk if you asked in the KER thread for a progress update on a specific feature or bug, as long as you asked in a decent manner.  The main KER git repository has an issue tracker though I don't know if there's one in there specific to this problem.

To answer the underlying question directly rather than the one you're asking here, no, little progress has been made with this so far.  I've been meaning to look into it but most of my recent KSP time has been spent on garbage related issues in the stock game and in other mods.  As far as I can see the code should be simulating correctly as long as the root part is one of the parts that doesn't get decoupled.  However, it appears to not be firing the next stage until all of the Xenon tanks on the vessel are empty, not just the ones that will be decoupled in the next stage.  I'll try to run some experiments over the next few days and see if I can spot the problem but no promises...

[Archgeek]

Drat and blast, sounds like I'm stuck in spreadsheet-town for a while longer.  Thanks for the answer, though.  I'm going to guess you've been cutting out LINQs and certain foreaches, just like the stock devs?  I hear those are big GC culprits.

[Padishar]

On 7/13/2016 at 6:22 PM, Archgeek said:

Drat and blast, sounds like I'm stuck in spreadsheet-town for a while longer.  Thanks for the answer, though.  I'm going to guess you've been cutting out LINQs and certain foreaches, just like the stock devs?  I hear those are big GC culprits.

Yes, I've been busy developing my MemGraph mod to both help diagnose garbage issues and to reduce the stutter, and also then helping to track down some of the major offenders in the stock game and in various mods.  KER also generates excessive garbage (though nothing like some others) but I've not started on addressing that yet.

In other news, I've taken a look at this Xenon droptank issue and it did turn out to be a very simple thing to fix.  I have submitted a pull request that fixes it and will post a test version of the latest KER code in the KER thread shortly.