Emmet Hikory wrote: > On Dec 18, 2007 9:09 AM, Thilo Six wrote: >> comparation of a whole install (download time + extract time): >> >> download time gz (39084/384)= 101.78s + 14.278s = 116.06s >> download time 7z (27358/384)= 71.24s + 143.783s = 215.02s > > Comparisons like this have far too many factors (as you mentioned > separately). I've seen Ubuntu used on a 3GHz Core2 Duo over ISDN and > on a 400MHz Pentium IIIs with ethernet connectivity to the mirrors, > which lead to compltely different results to the tests above. > > Bandwidth will always be location dependent, so package size > related to bandwidth is not an easy item for discussion. Smaller is > nice, but for some it doesn't matter very much, as their connections > are either fast enough that it's not important, or slow enough that > they download overnight (or otherwise deeply backgrounded) anyway. >
We should consider that it's not always the user's download that gets saturated, but the mirror's upload. This is especially true on release day. A 30% reduction in filesize would translate to a 30% reduction in download time for everyone. This becomes far more important when everyone's download time is over ten times as long as normal because of our mirror issues. More interestingly, it's quite likely that there are mirrors out there that become rate limited once a certain bandwidth threshhold is met due to cost concerns. If we prevent them from reaching that threshold by lowering the bandwidth, we could reduce download time by even more than the filesize reduction. Thanks, Scott Ritchie -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
