Handle return value 24 the same was as 23.

[merks]

eclipse-equinox/p2#166

[merks]

@umairsair

I'm not sure how to test if this is needed but it wold seem consistent with the exiting handling for 23.

[laeubi]

I just wanted to link this one here regarding exit codes it sometimes comes to a surprise:

• Suppress default behaviour of IApplication exit codes 23 and 24? #378

[github-actions]

Test Results

   27 files  +   25     27 suites  +25   11m 38s ⏱️ + 11m 27s
2 170 tests +2 156  2 124 ✅ +2 110  46 💤 +46  0 ❌ ±0 
2 228 runs  +2 200  2 182 ✅ +2 154  46 💤 +46  0 ❌ ±0 

Results for commit 8c65900. ± Comparison against base commit 850923a.

[merks]

This code has not been touched in a very long time so I doubt there is any one around who can tell us what's right or wrong. I think consistently handing 24 as we do 23 is most likely the correct and consistent thing to do and it's what I've done elsewhere to fix actual problems, e.g.,

eclipse-pde/eclipse.pde#1121

[HannesWell]

I wonder if there should be some global constants for that magic numbers?

[laeubi]

I wonder if there should be some global constants for that magic numbers?

You mean like

equinox/bundles/org.eclipse.equinox.app/src/org/eclipse/equinox/app/IApplication.java

Line 48 in 850923a

public static final Integer EXIT_RELAUNCH = Integer.valueOf(24);

equinox/bundles/org.eclipse.equinox.app/src/org/eclipse/equinox/app/IApplication.java

Line 39 in 850923a

public static final Integer EXIT_RESTART = Integer.valueOf(23);

:-)

In the end a real relaunch can only performed by a native launcher ...

[HannesWell]

I wonder if there should be some global constants for that magic numbers?

You mean like

For example, yes. :)
I think it would be good to use that where possible, i.e. platform and PDE. Here it is probably not accessible.

[merks]

I cancel this because eclipse-platform/eclipse.platform.ui#1622 avoids surfacing that exit code.