Gang bonus calculation issues when projecting boosting ship onto fit

[dachucky]

As per the subject, there seems to be something going wrong with the calculations when projecting A -> B when B is being boosted by A.

Examples:

Claymore with Shield Harmonizing Link projects onto unfit Retribution. Then set Claymore as Squad Booster. Notice that Retribution gets no bonuses.

Remove Claymore as booster and projected fit. Set Claymore as Squad Booster. Then project Claymore fit onto Retribution. Notice that Shield Harmonizing bonus is removed. Then remove Claymore from projected fits. Notice that all gang bonuses are removed, despite Claymore still being listed as Sqaud Booster.

[blitzmann]

Possible relation to #63

Haven't really had a chance to look into it yet, and it may be an inherit limitation with the current engine. Projections were hacked together and there are a lot of design flaws with them. I'm hoping against hope it's an easy fix, but...

Quick fix may just be to disallow projection of same ship as a booster, and vise versa. I cannot really think of a use case where the same ship would be projected and boosting at the same time...

[dachucky]

The use case for me was carriers and supercarriers with ganglinks and RR.

[DarkFenX]

As i mentioned multiple times, fleet boosting support is somewhat hackish. I noticed this bug several times and i suppose it is not an easy fix.

[blitzmann]

Popping in to confirm that this bug is still present as of v1.3.0. #63 was not the cause of this, but I did notice a similar issue when testing out fixes to that issue. Excerpt from pull request:

There is one issue related to this fix (assuming we use the fits in #63):

Attach the Sleipnir to the Scythe, then project the Scythe to the Oracle. Set the Sleipnir as the squad booster to the Oracle and notice that the stats do not change. Tab to another tab, then tab back. Stats show correct values. If you open the Oracle fit in this configuration, the correct values are also shown.

This shows us that it doesn't necessarily have to be the same ship being projected / boosting, but it does seem to happen if the same ship is in the chain of projections (if you have B boosting A, then have B > C > A projection). I have also confirmed that switching tabs in this issue recalculates the fit correctly (until you mess with the projections / boosts of the target ship).

I'll investigate this a bit more when I have time, and document some use cases here.

[blitzmann]

This should be fixed with latest commit in master (e74e261). If not, please reopen. =)