Re: on memory usage with the 64bit patch, and interpretation of various numbers

On Fri, May 16, 2014 at 9:50 AM, Dmitry Stogov <dmitry@zend.com> wrote: > Pierre, > > 1) if you reduce size off all language entities to 32-bit, then what this > patch would provide except of extending string to have 64-bit length? > but you always talk - it's a side effect (I'm not talking about IS_LONG > part). You keep using the same arguments in the loop and I keep answering the same thing as well. Not sure that will lead us anywhere. Anyway, let me try a last time. To be able to have 2^64 (-1 as zend_string use unsigned char) is by far not the goal but what is very well described in the links we posted, the answers we gave, etc. > 2) In phpng we use zend_string* for class, function and other names, so > extending zend_string would extend all of them anyway. Besides the fact that phpng is still a huge prototype and many things may still change in the next months/year, choices being done now does not mean they are written in stone. And as I said numerous times already, I did not have the time to analyze phpng more deeper yet while busy fixing it to make at least build or do not crash on supported platforms. Cheers, -- Pierre @pierrejoye | http://www.libgd.org