Member of the EVE Tweet Fleet

Monday, June 14, 2010

Shower: good for figuring things out...

Correction to the previous entry. I just figured out this morning that the problem is not with downtime. It's with the submit button in PI. Of course I figured this out IN the shower, right before I go to work so I can't test my theory right away.

The problem is whatever is causing all the cycles of a planet to synchronize. I still haven't been able to test it reliably but I just last night had a job where I put in exactly 1 cycle worth of P1 and got zero product. I suspect it has to do with the submit button resetting things or at some point after the submit the synchronization happens which causes existing production to blow up and new raw materials to be sucked into the processor.

More as I test this later tonight...

Man this new PI is buggier than a 3 month old corpse.

It also shows the way towards a workaround. I'll need to synchronize by having any route changes and res-scanning of resources all happen at the same time in a single submit. That might be the solution to avoiding the losses of materials. Need to test...

7 comments:

Toldain said...

You realize, don't you, that each industry dot has storage capability equal to the inputs for whatever schematic is selected? And that this is above and beyond the job in progress. It's what lights up the little squares when you open it, and the lights up the border of the dot, with bright orange.

I bet that if you change the schematic before running those dry, those materials get dropped on the floor.

This is only an issue if you are timeslicing, which I haven't done yet. Well, I did on SISI, and I seem to recall losing those materials.

Letrange said...

Yep. But the materials are also thrown away if you don't do anything and change something else then hit submit. So for example say you get your production setup. Then hit submit. Then you go and scan your resources to get them going again 3 hours later. That seems to be blowing up the production that you did not touch. (that's the current working hypothesis atm)

Benoit CozmikR5 Gauthier said...

Let's start taking bets on which part of PI the vikings will leave unfinished...

Letrange said...

Which part? you mean how many parts don't you?

Benoit CozmikR5 Gauthier said...

Short list of unfinished parts: missing 5th T3 subsystem, T3 frigs, missing assault frig bonus, rocket fix, corp UI... did they fix fleet fights?

I don't know which part of PI, but this seems like a rocky start.

Anonymous said...

OK. I am trying to figure out your problem, but we seem to be making things a bit differently.

All my resources funnel into and out of the launchpad. Processors are always on, so when there is enough material in the launchpad, a processor sucks it in and starts processing, placing the new material back to the launchpad.

I take it you are running different schematics each time? So, if there is material already queued into the proc, you then changed schematics, and the first material goes poof? I assume that is what is happening??

Letrange said...

@Zuraku, that's correct I'm switching stuff. BUT I'm waiting for the existing input to run dry before changing anything. It's definitely a little in-efficient but it permits me to manufacture all P4 sequentially if not simultaneously.