[8.x] Revert primary key stuff

[driesvints]

This reverts #37715 & #37782.

Fixes #37820 #37811 #37781

[khalilst]

@driesvints With all due respect, I think you are in a hurry to revert.
The bugs related to that PR could be easily fixed.

Please consider this issue #33238 and @taylorotwell invited to solve that problem.

[khalilst]

@taylorotwell
This bug #37820 related to the third-party package overriding the Blueprint.

The package name: bosnadev/database
File path: Bosnadev/Database/Schema/Blueprint.php
It is required by mstaack/laravel-postgis.

Also, there is an issue and PR in that package to solve this problem.
bosnadev/database#48
bosnadev/database#49

The two other bugs are fixed with tiny changes.
I think reverting these PRs #37715 & #37782 is not fair.

[Shkeats]

I had to lock back to 8.47.0 to avoid SQLSTATE[42000]: Syntax error or access violation: 1068 Multiple primary key defined with our existing mysql migrations. I think it's right to revert this stuff as its a breaking change.

[khalilst]

@Shkeats , If you would share the failed migration, I may fix it.

[Shkeats]

@khalilst I don't think it's really a scalable approach for you to rewrite everyone's migrations, although the offer is appreciated!

[Shkeats]

We don't have bosnadev/database in our codebase and I suspect my issue is already covered by the other ones reported. However, if it's helpful at all @driesvints I can open a new issue with our specific case.

[driesvints]

@Shkeats can you try the new release we just made?

[GrahamCampbell]

🚢

[Shkeats]

@driesvints I'm still seeing 8.48.1 as latest from what I can tell? I tried applying this change locally but it didn't help in my case.

[taylorotwell]

@khalilst we have to revert this. We have more people in this thread reporting breakages that aren't related to a package. Laravel is used by many many thousands of applications and I can't just leave them broken. Sorry.

[taylorotwell]

@khalilst would be willing to explore this again at a later time but it clearly needs more thorough exploration of what these breaking changes are.

[khalilst]

@taylorotwell I'm just trying to fix the issues, but as you can see here @Shkeats even did not provide enough information and I think it was not a good idea to consider it as a breaking change!

BWT, Thank you so much and I think you have good reasons to do this.

[khalilst]

@khalilst would be willing to explore this again at a later time but it clearly needs more thorough exploration of what these breaking changes are.

@taylorotwell Sure, I will create a package and after resolving all bugs, I will create another PR for this thread.