Development: Drop the unused result and build log order columns

[krusche]

Summary

Drops result.results_order and build_log_entry.build_log_entries_order. Neither column has a reader any more: the correction round moved onto the result and the build log entries are ordered by the timestamp they already carry.

Important

Do not merge before the release that contains #13570 has shipped. That PR stopped using both columns but deliberately left them in the database, so that rolling the application back to the previous version would still work. Dropping them in the same release would take that away. This is a draft until then.

It is also stacked on feature/programming/minimize-result-processing-data, so the base has to be switched to develop once #13570 is merged.

Motivation and Context

Follow-up cleanup for #13570. Both columns existed to let Hibernate maintain the position of an entry inside an @OrderColumn collection. Maintaining that position is what forced every write in those areas to go through a save of the whole submission, and saving a submission selects it together with its participation, its exercise and the exercise's course, because all three associations are eager. That is where most of the ~140 KB per build result came from.

Now that the position is not needed, the columns are only occupying space and inviting someone to rely on them again.

Description

One changelog with two changesets, each guarded by a columnExists precondition with onFail="MARK_RAN", so the migration is idempotent and does not fail on a database where the column is already gone.

No code changes: neither column is mapped or referenced anywhere.

Steps for Testing

Prerequisites:

• 1 Instructor, 1 Student
• 1 Programming Exercise on a server with the integrated lifecycle setup (LocalVC and LocalCI)

1. Upgrade a server that already runs the release containing Development: Minimize data transfer when processing a build result #13570. The migration applies without error.
2. As the student, push a passing commit and then one that does not compile. Both produce a result, the failing one is marked as build failed, and its build logs are readable and in chronological order.
3. As a tutor, open the exercise for assessment and check that the result belongs to correction round 0.
4. On an exam exercise with two correction rounds, assess both rounds and confirm each round shows its own result.

Server Tests

762 server tests were run locally with both columns dropped, across the result-processing, assessment, submission-policy, localci and participation suites. Since the test schema is built by Liquibase, this exercises the drop itself: if anything still expected either column, these would fail.

Review Progress

Code Review

• Code Review 1
• Code Review 2

Checklist

General

• I tested all changes and their related features with all corresponding user types on a test server.
• Language: I followed the guidelines for inclusive, diversity-sensitive, and appreciative language.
• I chose a title conforming to the naming conventions for pull requests.

Server

• Important: I implemented the changes with a very good performance and prevented too many (unnecessary) and too complex database calls.
• I strictly followed the principle of data economy for all database calls.
• I strictly followed the server coding and design guidelines and the REST API guidelines.

Changes affecting Programming Exercises

• High priority: I tested all changes and their related features with all corresponding user types on a test server configured with the integrated lifecycle setup (LocalVC and LocalCI).

[github-actions]

End-to-End Test Results

Phase	Status	Details
All Tests	❌ Failed	Tests Passed ☑️ Skipped ⚠️ Failed ❌️ Time ⏱ All E2E Tests Report (PR) 376 ran 367 passed 8 skipped 1 failed 37m 20s

❌ Failed Tests

• Exam assessment › Modeling exercise assessment › Instructor makes a second round of assessment (6m 11s)

Test Strategy: Running all tests (configuration or infrastructure changes detected)

Overall: ❌ E2E: real (non-flaky) test failure

🔗 Workflow Run · 📊 Test Report