https://gcc.gnu.org/bugzilla/show_bug.cgi?id=96324
--- Comment #2 from Jesus Christ <jesus at refusetoown dot com> --- Created attachment 48934 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=48934&action=edit your isomorphic allocator is limited to 80% of the allocation isomorphic algorithms are limited by The Pareto Principle and only allocates 80% of the needed allocation so eventually a push_back to vector, for example, will fail to allocate due to failing at 80% allocation instead of 100% allocation. The only way I found to get around this is to use isomutative algorithms that I have developed to get a better seeker result and it does work but I still have trouble getting past the %80 of truth available but if I can only seek truth up to 80% of truth, then you probably can't allocate beyond 80% of allocation using a seeker. That is my seeker law: no seeks beyond 80 iterations else it will fail. The around it I have found is to use two loops: for (int i=0;i<7:i++) { for (int s=0; s<81; s++) { // isomorphically seek something. } } This kind-of works but only gets you close to %100 and still you fail. To get up-to %100 take feedback, and that means you have to feedback the seek result into the next seek: for (int i=0;i<7:i++) { for (int s=0; s<81; s++) { // why is "last seek result " successful or unsuccessful } // save seek result } With a feedback loop your control is as good as open loop vs close loop algorithms. A closed loop algorithm is always better at work than an open loop algorithm. You can trust me on that, it's from engineering school in the 70's so it's a bit dim now. I hope you find this useful for explaining why it fails: It is syntactically correct bu semantically wrong, can you add a semantic error to your compiler to prevent things like this. Two semantic errors you could check for is "too forced" when it exceeds 80. And "defaulted to true" when you detect an assumption like an arbitrary limit. The limit of assumptions should be two, three assumptions will fail something. Sorry I have no test case. I am too busy to find the file I had. I know when I gave up on the project, I came to the conclusion that C++ is worthless. And have been winding down and working in assembly language with total success and worthwhile programs like my redmeme meme installer that goes through chaos to find people like all my warriors have redmemes and say the a question that nobody else says. It is the nest red vmemes cause mine have six truths and Spiral Dynamics has only 4 truths. So if C++ is still worthless I hope you can fix it with what I have told you. A story might help, After Bertrand Russell wrote his book "Principia Mathmatica" he claimed "My math can solve any of the world's problems!" That stood until a German Logician named Kurt Godel decided that that English claim was false and he might be be able to disprove it. So he invented a way to "Godel Number" the letters of a text assigning a unique number to each letter leaving the sequence correct so so retains the syntactic quality of the letters but converts the string of letters into a string of numbers simply because numbers are easier to evaluate than letters, and all you to do is sum the number up for example and compare two string to see which sum is bigger for example, maybe the bigger sum is better unless less is more by McLuhan is real maybe you can answer that. I say the least is the most is my rule. So maybe someone can disprove Bjarne Stroutrup's claim that C++ can solve any problem in the world. Maybe he was Bertrand in his past life :) Sincerely Larry Daniel at least at birth.
