[jRaven]

trunk/plugins/include/sqlx.inc

stock Handle:SQL_MakeStdTuple()
{
    static host[64], user[32], pass[32], db[128]
    static get_type[12], set_type[12]
    
    get_cvar_string("amx_sql_host", host, 63)
    get_cvar_string("amx_sql_user", user, 31)
    get_cvar_string("amx_sql_pass", pass, 31)
    get_cvar_string("amx_sql_type", set_type, 11)
    get_cvar_string("amx_sql_db", db, 127)
    
    SQL_GetAffinity(get_type, 12)
    
    if (!equali(get_type, set_type))
    {
        if (!SQL_SetAffinity(set_type))
        {
            log_amx("Failed to set affinity from %s to %s.", get_type, set_type)
        }
    }
    
    return SQL_MakeDbTuple(host, user, pass, db)
}

Shouldn't the following pass a value of sizeof(get_type)-1 = 11 instead of 12?

    SQL_GetAffinity(get_type, 12)

Is there any performance disadvantage to using sizeof directives in stocks like this? It seems like it would be a better or safer practice to have declaration-relative maxlens automatically put in place by the machine.

[MaximusBrood]

First of all, I'm not sure if I'm making sense here.
I just wanted to say that sizeof will not affect performance, it's handled at compiletime, not at runtime.

[Zenith77]

It should be 11, however using sizeof really wouldn't make sense since you know the size.

[BAILOPAN]

This is indeed a typo. sizeof() is encouraged on local variables whose size is pre-defined, since you don't have to update constants throughout code.

[BAILOPAN]

Thanks for the report. This is fixed as revision 3174.