Giorgos Keramidas wrote: > > Compression gets rid of about 36MB. > > How long did that take to compress though?
2 minutes on a P3-800 with 128M of RAM and one IDE disk. Doesn't matter, because all it really adds is latency. > What load did the machine that did the compression have? Currently, > the snapshots.jp.freebsd.org machines build releases every 24 hours > which last 4-9 hours. I'm not sure if the same machines could spare > some cycles to compress the ISO images, or the disk space to store > almost duplicate copies of the same ISO images twice (compressed, and > uncompressed). The best people to ask about things like these are > the jp.freebsd.org admins and not a local compression program imho. The space argument may be valid; though, in that case, you'd expect that compressed images would be the only images that would be there. 8-). > > I think the correct answer is maybe "because the FAQ maintainers > > have broadband connections"... > > No we don't. My "ultrafast" connection is in fact a 28.8 Kbit/sec > dialup connection. This is why I don't download entire ISO images, > but instead do FTP-installs. So, there you go ;) It's *incredibly* hard to get a -current machine initially installed from sources corectly. It's easier to use the ISO's, even if they take a very long time to download. It's either that, or don't start following -current. > Not very irrelevant, as it might seem at first. Because I'm not > talking about the FTP server that delivers the files, but about the > server that 'builds the snapshots'. > > The donations list of freebsd.org lists requests for better, faster > release building machines for the Japan cluster. If you really think > that you can help, I'd be glad to be proven wrong by a generous > donation to the guys who have saved my -current installation at home a > dozen times with their snapshots. 3.5 hours worth of additional FTP downloading time per download, vs. two minutes of compression time... -- Terry To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message