Use the index 2 while benchmarking, or else the std::visit will fail, which
again proves my point on the fact that "std::visit is just a dumb union"

regards, Muneem

On Thu, Apr 2, 2026 at 9:46 AM Muneem <itfllow123_at_[hidden]> wrote:

> basically, the turn of events were these:
> dealing with the errors in https://quick-bench.com/ was painful because
> it wasn't accepting user input when i debugged it all, then when i complete
> the program, I didn't have a compiler, then had a emotional fight with my
> mom, then tried installing a compiler, failed at first, installed an old
> one at the second, then at last, it worked. Sorry for trauma dumping, but I
> just cant hold it back when I am missing on time with the math book that I
> love so dearly(my routine didn't turn out the way it was supposed to), and
> during all that I got an email from a guy from the standard(according to
> linkdin) calling my idea stupid and to not continue it; Life really had to
> give me this meltdown. Sorry for the trauma dump story. Ignore it, please
> (for me)
>
> On Thu, Apr 2, 2026 at 9:45 AM Muneem <itfllow123_at_[hidden]> wrote:
>
>> Hi!
>> //sorry for my late bench mark results; I had a very unfortunate turn of
>> events in the past few hours. REALLY REALLY SORRY!!!!
>> #include <variant>!!
>> #include <iostream>
>> #include <chrono>
>> #include <ctime>
>> #include <iomanip>
>> struct A { inline int get(){return 1;} };
>> struct B { inline int get(){return 2;} };
>> struct C { inline int get(){return 3;} };
>>
>> struct data_accessed_through_visit {
>> static std::variant<A, B, C> obj;
>> int operator()(int) {
>> int a1;
>> std::visit([&a1](auto&& arg) { a1= arg.get(); }, obj);
>> return a1;
>> }
>> };
>>
>> std::variant<A, B, C> data_accessed_through_visit::obj = C{};
>> int a_val = 1, b_val = 2, c_val = 10;
>> int user_index = 0;
>>
>> struct data_ternary {
>> int operator()(int index) {
>> return (index == 0) ? a_val : (index == 1) ? b_val : (index == 2)
>> ? c_val : -1;
>> }
>> };
>>
>> struct data_switched {
>> int operator()(int index) {
>> switch(index) {
>> case 0: return a_val;
>> case 1: return b_val;
>> case 2: return c_val;
>> default: return -1;
>> }
>> }
>> };
>>
>>
>>
>>
>>
>> volatile int x = 0;
>> constexpr uint64_t loop_count=100000;
>> static void measure_switch() {
>> data_switched obj;
>> for (int i=0; i++<loop_count;) {
>> x = obj(user_index);
>> }
>> }
>>
>> static void measure_visit() {
>> data_accessed_through_visit obj;
>> for (int i=0; i++<loop_count;) {
>> x = obj(user_index);
>> }
>> }
>>
>> static void measure_ternary() {
>> data_ternary obj;
>> for (int i=0; i++<loop_count;) {
>> x = obj(user_index);
>> }
>> }
>>
>> template<typename func_t>
>> void call_func(func_t callable_obj, int arg){
>> const auto start = std::chrono::steady_clock::now();
>> callable_obj();
>> const auto end = std::chrono::steady_clock::now();
>>
>> std::cout<<
>> std::chrono::duration_cast<std::chrono::nanoseconds>(end -
>> start).count()
>> <<std::endl;
>>
>> }
>>
>> int main(){
>> std::cout << "Enter index (0 for A, 1 for B, 2 for C): ";
>> if (!(std::cin >> user_index)) std::cerr<<"error!";
>> if (user_index == 0) data_accessed_through_visit::obj = A{};
>> else if (user_index == 1) data_accessed_through_visit::obj = B{};
>> else if (user_index == 2) data_accessed_through_visit::obj = C{};
>> std::cout<<"time taken for switch statements:";
>> measure_switch();
>> std::cout<<"time taken for visit:";
>> measure_visit();
>> std::cout<<"time taken for ternary:";
>>
>> measure_ternary();
>> return 0;
>>
>> }
>> /*
>> results:
>> Time (ns) for switch: 442900
>> Time (ns) for visit: 3682600
>> Time (ns) for ternary: 248300
>> Told you!!!
>> thank you for all of you're feedbacks!!!
>> */
>> regards, Muneem
>>
>>
>> On Thu, Apr 2, 2026 at 8:26 AM Muneem <itfllow123_at_[hidden]> wrote:
>>
>>> The benchmarking tools that you (Breno Guimarães )
>>> suggested didn't work because they do compile time optimizations on the
>>> indexing itself (if the index can be determined at compile time), and when
>>> I tried to make them read the index, it didn't work as expected, hence it
>>> will take some time for me to write my own benchmarking code..
>>> To the response I got from a very senior person, I wanna say that even
>>> though this idea might not be within the logical framework of C++, I would
>>> still push for it until there is a mathematical proof to specify otherwise.
>>> I hope this isnt disrespect. Sorry for the delay on benchmarks too, guys; I
>>> was busy eating breakfast.
>>>
>>> On Thu, Apr 2, 2026 at 8:24 AM Muneem <itfllow123_at_[hidden]> wrote:
>>>
>>>> Before I add some code, I wanna specify that I know references are the
>>>> same as pointers,but compilers can (in theory optimize them better because
>>>> they are like constant pointers).
>>>>
>>>> On Thu, Apr 2, 2026 at 6:35 AM Breno Guimarães <brenorg_at_[hidden]>
>>>> wrote:
>>>>
>>>>> Try adding actual code showing the problem and how it looks like with
>>>>> the different options you claim de be not sufficient and how the code would
>>>>> look like with your solutions.
>>>>>
>>>>> There are many wrong assumptions on how things work today. I honestly
>>>>> cannot understand what you are actually proposing.
>>>>>
>>>>> Em qua., 1 de abr. de 2026, 20:37, Muneem via Std-Proposals <
>>>>> std-proposals_at_[hidden]> escreveu:
>>>>>
>>>>>> This is the updated proposal
>>>>>>
>>>>>> On Thu, Apr 2, 2026 at 1:45 AM Muneem <itfllow123_at_[hidden]> wrote:
>>>>>>
>>>>>>> Ok, I will draft a format proposal that outline all the problems (a
>>>>>>> whole class of problems), cite all the possible solutions, and then cite
>>>>>>> bjarne Stroustrup quotes from his old 2013 book to back my philosophy up.
>>>>>>> Give me some time and thank you for your feedback ❤️❤️
>>>>>>>
>>>>>>> Regards, Muneem
>>>>>>>
>>>>>>> On Thu, 2 Apr 2026, 1:15 am Thiago Macieira via Std-Proposals, <
>>>>>>> std-proposals_at_[hidden]> wrote:
>>>>>>>
>>>>>>>> On Wednesday, 1 April 2026 10:52:08 Pacific Daylight Time Muneem
>>>>>>>> via Std-
>>>>>>>> Proposals wrote:
>>>>>>>> > Okay, I won't make the JIT part in my proposal, but just so I
>>>>>>>> know, you do
>>>>>>>> > support me if I dont? Like you will vouch for me and help me in
>>>>>>>> my actual
>>>>>>>> > formal proposal?
>>>>>>>>
>>>>>>>> I haven't understood the problem yet. You need to start the
>>>>>>>> proposal with the
>>>>>>>> problems (plural) that need solving, preferably with multiple
>>>>>>>> possible
>>>>>>>> alternatives and showing why you think your solution is best.
>>>>>>>>
>>>>>>>> --
>>>>>>>> Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
>>>>>>>> Principal Engineer - Intel Data Center - Platform & Sys. Eng.
>>>>>>>> --
>>>>>>>> Std-Proposals mailing list
>>>>>>>> Std-Proposals_at_[hidden]
>>>>>>>> https://lists.isocpp.org/mailman/listinfo.cgi/std-proposals
>>>>>>>>
>>>>>>> --
>>>>>> Std-Proposals mailing list
>>>>>> Std-Proposals_at_[hidden]
>>>>>> https://lists.isocpp.org/mailman/listinfo.cgi/std-proposals
>>>>>>
>>>>>