Reply With Quote
i think that it is bad policy - writing compressor for testers instead of users. in particcular, i think that most well-known tests don't reflect actual data comopressed by users - large folders of spurces, documents, executables. these strange tests then goes to the user's requirements to add MM compression or something else that they will not really use
imho, testing of gimp sources, windows executables or html docs is the best way to understand which programs are good and which are bad for real users. Werner though about adding such test but not done it
i can propose to all tests authors make, for example, test of OpenOffice sources, executables (and html docs, if possible) like it was done by Squeez authors; and include comp/decomp time here because otherwise paq don't have any competitors
othersie, i personally will continue to mainly ignore such tests because they, imho, don't reflect compression of actual data in real life
Results 1 to 30 of 39
Originally Posted by Black_Fox
