Le 17 nov. 2012 à 10:15, Christian Schmitz a écrit: > I'd simply use zlib with level 9. > But it'S not wrong to have a flag in your data which says what mode you use > like 0 for not compressed, 1 for zlib and 2 for bzip and 3 for lzw.
Ok, thanks. since yesterday, I've added a parameter (argument of my console app) so I could specify the compression (this value is stored in the file, of course). Ideally, I'd like to avoid “bad” compression methods which would sometimes fail (like one with a reputation of not always decompressing right). Are there unreliable compressions in these above? Or is it always more risked to use compression? > So your app could see which one is the smallest and store that. Sometimes > uncompressed is smallest :-) Really? I didn't expected that. What I could do is have 4 variables in code, one for each compression (and for no compression) and always save the one that is the smallest. But that would fill the RAM quickly (when copying big files). Ok, at least I'm going to compare the file sizes. Thank you. _______________________________________________ Mbsplugins_monkeybreadsoftware.info mailing list [email protected] https://ml01.ispgateway.de/mailman/listinfo/mbsplugins_monkeybreadsoftware.info
