On 02/21/2013 09:29 PM, Julius Davies wrote: > Hi, > > I would like to go ahead with my original patch, or at least > understand better how I can improve it, or why it's not acceptable. > > Thomas did manage to improve our numbers with a very nice & small > patch (here's a benchmark comparison): > > https://issues.apache.org/jira/browse/CODEC-166?focusedCommentId=13581834&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13581834 > > But I think it's still worth it to take this patch as well, since it > still boosts things another 75% on encode and 20% on decode on top of > Thomas's improvements. > > I'm not in any rush, but I thought I'd post this to start a new thread. > > > References: > https://issues.apache.org/jira/browse/CODEC-166 > > Original thread to (reverted) commit: > http://markmail.org/message/xvb6nzfdlthzjcnu
as nobody responded so far: Having the (probably) fastest java Base64 implementation in Codec would definitely be nice, we just need to make sure it does the same thing as the existing implementation (considering all the options you can provide). I did not like the fact that there was another source file, providing another Base64 implementation (actually its already the third one in codec, as there is another slightly different one in the digest package). Thomas --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org
