[shauli]

Originally Posted by ^SmileY I know that g_Stats[i][Objective] is same as g_Stats[i][4]

g_Stats[i][Objective] will be MAX_AUTHID_LENGTH+MAX_NAME_LENGTH+STATSX_MAX_ STATS+MAX_BODYHITS, something like 83 or so, not 4. So the compiler reads it as 83 + STATSX_MAX_OBJECTIVE instead of just STATSX_MAX_OBJECTIVE.
That's why you get an error.
It will also happen with get_players() or any function that expects a fixed size. I believe it is not your problem and that it should be fixed somehow in the new AMXX version.

[Bugsy]

Originally Posted by ^SmileY get_stats2 have a fixed array for parameter. that is why im having troubles. The solution is use as separated array, or just use a copy() to copy into the struct.

copy() only works on strings, not arrays. So it could work for ascii values (0-255), but if items 0, 1, or 2 are equal to 0, the data copy will end there. Not reliable for this purpose. Sadly, I don't think there is an array copy native.

Originally Posted by shauli g_Stats[i][Objective] will be MAX_AUTHID_LENGTH+MAX_NAME_LENGTH+STATSX_MAX_ STATS+MAX_BODYHITS, something like 83 or so, not 4. So the compiler reads it as 83 + STATSX_MAX_OBJECTIVE instead of just STATSX_MAX_OBJECTIVE. That's why you get an error. It will also happen with get_players() or any function that expects a fixed size. I believe it is not your problem and that it should be fixed somehow in the new AMXX version.

No it doesn't, it will be sized at 4. Check for yourself : sizeof( g_Stats[][Objective] )

This should technically work, it just doesn't.

[shauli]

Originally Posted by Bugsy No it doesn't, it will be sized at 4. Check for yourself : sizeof( g_Stats[][Objective] ) This should technically work, it just doesn't.

I was talking about the index, not the size.
g_Stats[i][4] will return the 5th char in Auth, g_Stats[i][83] (or something like that) will be Objective.

[Bugsy]

Correct, the elements of the enum sized array are packed side by side, so from that perspective yes. But the compiler should handle this scenario.

[klippy]

This actually still exists in SourcePawn as well, and Dvander says it's not an easy fix: https://github.com/alliedmodders/sourcepawn/issues/301