Closed Lisias closed 1 year ago
See #64 for logs, crafts and more information about the problem
Well, I found the reason this thing is screwing up the user.
What I don't understand is why in fscking hell this piece of krap is working fine on my rig.
Commit https://github.com/net-lisias-ksp/KSP-Recall/commit/c198a1389d8db409c96061c23250c24bfbdecff3 restores the feature.
Commit https://github.com/net-lisias-ksp/KSP-Recall/commit/a08690d0f0ed21543b60528d8012db2e5d1b96e5 does a salvage on current SFS and CRAFT files affected by this huge bork of mine.
Now I need to understand why in hell I didn't detected this on my Acceptance tests.
JESUS FSCKING CHRIST. Found it.
IT'S MODULE MANAGER. This piece of crap fails now and then from detecting changes on the GameData, and fails to update the Config Cache.
So I had tested a bad code using an old Config Cache, because the fscking Module Manager didn't detected a change on the DLLs!!!
By manually deleting the ConfigCache, my new crafts started to have the same problem I detected on the Krazy1's.
I was a sinner, but I had seen the Light and now I'm a believer - because only God Himself would be able to keep this Community afloat for so many years with this abhorrent piece of crap being used as the most important component of the modding scene!!!
The solution? Easy. I'm ditching Module Manager for good now, even on the Acceptance tests.
From now on, I will only use my own fork, and anyone asking for support will be advised to consider switching to it. My fork have this problem tackled down.
There's something additional happening - I eye balled the MM (Forum) source code, and didn't found any new weakness that I had not detected before.
DIGESTs are not IDENTITIES, and MM (Forum) fails bitterly on recognising this fact, so evidences still pinpoint it as the source of the problem - but I will need some more digging until I'll able to pinpoint the exact piece of code that it's failing.
An issue on MM/L was created, related to this happening.
POST MORTEM:
About my comment on : https://github.com/net-lisias-ksp/KSP-Recall/issues/65#issuecomment-1500777219
Well… This time MM wasn't the responsible for the problem, it was only involved.
I created a bug on KSP-Recall where the affected PartModule
was not being automatically activated on creation. I didn't detected the problem because I was reusing Crafts and Savegames already existent, where the PartModule
s were already activated, and so nothing wrong happens.
But on new crafts and savegames, the affected PartModule
was being injected deactivated.
My MM fork mitigated the ConfigCache problem, but in theory had not really solved it - only make it absurdly improbable to happen, so anything that woks on my fork and doesn't on Forum MM I tend to automatically pinpoint Forum's MM as the culprit.
BUT THIS TIME, I got screwed by myself because while testing and developing the last Recall release, I was reusing crafts and savegames and so didn't noticed the problem. On the acp
(acceptance) test bed, I created a new savagame with new crafts and, so, I got bitten by my mistake. But since the acp
test bed uses Forum's MM, I misdiagnosed the source of the problem - and as we can easily note, I'm absolutely fed up with Forum's MM (and some other add'ons) stupidities, and this is interfering with my diagnosing sometimes.
(sigh).
Have you ever stopped to consider that maybe you're just not always right and that jumping to conclusions and calling other people's code trash makes you kind of a dick?
Have you ever stopped to consider that maybe you're just not always right and that jumping to conclusions and calling other people's code trash makes you kind of a dick?
I'm a dick only when I wrong, my "friend". :D
By the way:
Discussed in https://github.com/net-lisias-ksp/KSP-Recall/discussions/64