Closed Bug 412732 Opened 17 years ago Closed 17 years ago

compatibility.ini reports a bogus LastOSABI on Intel

Categories

(Toolkit :: Startup and Profile System, defect)

x86
macOS
defect
Not set
normal

Tracking

()

VERIFIED INVALID

People

(Reporter: alqahira, Unassigned)

Details

I noticed a few weeks ago when I was unable to install Benjamin's Memory Corrupter xpi (to test Ted's first Crash Reporter rework) that I was getting a strange build type:

"'Memory Corrupter' could not be installed because it is not compatible with your Minefield build type (Darwin_x86-gcc3).  Please contact the author of this item about the problem."

Of course, we can't build an x86-gcc3 binary on Mac OS X.  

I noticed tonight while looking at something else that my Camino profile's compatibility.ini declared this same bogus value for its LastOSABI, and I then checked my other profiles, branch and trunk, Camino and Firefox, and they all report this same bogus value (there is nearly-identical code in toolkit/xre/nsAppRunner.cpp and camino/src/preferences/PreferenceManager.mm that presumably is causing the bogus values; I can file a second bug against the Camino code tomorrow) on Intel.

My PPC is still running 10.3.9, so I don't know if the trunk PPC value is also wrong; branch PPC is reported sanely.
gcc3 is the ABI, not the actual compiler, gcc3 is correct.  FWIW, the reason you saw that was because memory-corrupter.xpi's install.rdf has this line:
<em:targetPlatform>Darwin_x86-msvc</em:targetPlatform>
Status: NEW → RESOLVED
Closed: 17 years ago
Resolution: --- → INVALID
Ah, OK.  Thanks for straightening me out :)
Status: RESOLVED → VERIFIED
Component: XRE Startup → Startup and Profile System
QA Contact: xre.startup → startup
You need to log in before you can comment on or make changes to this bug.